".
.
~
International Council for
the Exploration of tne Sea
ICES C.M. 1992/G:45
Demersal Fish Committee
IDENTIFICATION OF DANISH NORTH SEA TRAWL FLEETS AND FISHERIES
•
DANISH INSTITUTE FOR FISHERIES AND MARINE RESEARCH
CHARLOTTENLUND CASTLE
DK-2920 CHARLOTTENLUND
ABSTRACT
•
Many North Sea fisheries are mixed-species fisheries, which make
management of technical interactions problems important. This
kind of managemen~ requires that the main fisheries are identified and described with respect to species composition and
spatial and temporal, distribution of landings. Furthermore,
appropriate identification of main fisheries and the corresponding catch and effort data may lead to more useful estimates of
catchability for tuning of Virtual Population Analysis (VPA) or
integratedstochastic models.
Cluster analysis and principal component analysis have been used
as a general method of classification. Two different approaches
have been taken. Firstly, classification of individual trips
gives an overall desqription of the Danish trawl fishery and may
lead to 'more accurate catchability estimates due to' the close
conne'ction of effort and target species. Secondly, classification
of annual landings by vessel results in manageable subfleet
divisions by metiers~.
1
.
~
Introduction
Overfishing of some demersal North Sea stocks is one of the
reasons for the in'creased need for more refined and detailed
management measures. Most North Sea stocks are presently
regulated using TAe's, minimum mesh size and minimum landings
size as the dominating management instruments. However, as many
roundfish fisheries have been mixed fisheries, these instruments
seem to insufficient. In this connection management consideration
taking technical interaction into account has become urgent.
•
Dealing with technical interactions requires, contrary to Total
Allowable Catch (TAC), that total catch should be divided into
fleets (e.g. by gear and nation) for which the species compositions are known. Multi fleet prediction models taking technical
interactions into account are in principle models for which
partial fishery mortality rate by species and fleet are treated
separately for each fleet but simultaneously for all species.
This type of age s~ructured models are used in many areas, for
example the STCF model considering the North Sea (Anon, 1991 a),
The Celtic Sea treated by (Anon, 1991 b) and (Laurec et al 1991)
and the Gulf of Maine by (Murawski et al 1991). Other models
perform both age and length structured multi-fleet and multispecies predictions (Mesnil andShepherd 1990).
When performing multi fleet predictions using above mentioned
type of models it. is implicitly assumed, that the species
composition of a fleet remains .constant for the prediction
period. However, if managers want to change the rate of exploitation of specified target species/fisheries or if multi-species
TAC's should incorporate the effect of technical interactions,
then a division into fleets is not sufficient for evaluation of
such changes. To do that it is necessary to identify "metiers" or·
main fisheries by fleet and obtain estimates of the corresponding
effort and species composition of the catch.
•
Identification of maln fisheries may be trivial if all species
are caught in weIl separated "clean" fisheries. This is not the
case for many North Sea fisheries, where many species are caught
in the same areas at the same time. This applies especially to
the roundfish speci~s, cod, haddock, whiting and saithe, and to
some extent to some flatfish species as weIl.
Identification of main fisheries in the North Sea has until yet
not been carried out on a scientific basis and generally only
rather limited work has been done in this area. As far'as we can
see only.the Americans and the French have previously dealt with
this question. Main fisheries may be defined and identified in
several ways. (Murawski et al. 1983) analyses the otter-trawl
fisheries off the northeast coast of Uni ted States and define
fisheries by grouping categories of areas, depth zones and time
periods. They used cluster analysis technique as an objective
met):lod of fisheries identification. Besides of management of
individual fisheries this kind of definition is especially
sui table for management measures as box closures in time and
space ..
2
(Biseau and Gondeaux 1988) and (Laurec et al. 1990) defines
"metiers" of French Cel tic fleets as a combination of gear,
target species and area using principal component analysis.
This paper present analyses of Danish North Sea trawl fisheries,
which includes small mesh fisheries for industrial purposes and
human consumption fisheries. As some fisheries very often are
mixed-species fisheries because of spatial co-occurrence of the
species it is in this cases not a satisfactory solution to define
main fisheries on an area basis as above. Instead main fisheries
are defined as objective classifications of the observed species
composition (target species) of the landings, where any relevant
time scale could be considered.
•
•
Such classifications may be useful for management of individual
species involved in mixed-species fisheries e. g. when identifying fleets fishing for cod, which has earlier been tried. If the
spatial distribution of the landings or the fisheries are known
the effects of box closures in time and space may be evaluated as
weIl .
Identification of fleet fisheries may be carried out in several
ways depending on the purpose of the analyses. Two purposes are
considered:
A.
To obtain main fisheries, based on as disaggregated
catch information as possible, for which the corresponding effort is closely related to the catch of
specified target species.
B.
To identify operational sub-fleets homogenous with
respect to species composition of catch. These subfleets grouped by metiers could be the basis for
management of mixed-species fisheries taking technical interactions into account.
The first purpose could in principle include catch analyses by
each haul. However, as catch information by vessel trip is the
most detailed levelof aggregation available, this will·form the
basis of the analysis. This may lead to improved estimates of
catchability by sp~~ies for tuning of VPAs or integrated
stochastic methods.
This analysis of the North Sea trips is not a classification of
the vessels because each vessel may contained in all or many of
the trip classificatio~s.
Therefore, the second purpose deals with an operational classification of the vessels, which could be used for management
purposes . This could be achieved applying monthly or annual
landings of the vessels.
3
.
-
Data material
The present analyses are based on a comprehensive database for
1988 describing the Danish North Sea fishery by each individual
trip. Data stern from four main sources: Sales slip, a vessel
register, log books and biological data. The database contains
the following information by each vessel trip:
Vessel size category
Gear and mesh size
Year, month and date
ICES statistical rectangles
Landing category (human consumption or industrial fishery)
Landings in weight and value by species
Days absent
654 trawlers greater than 10 GT and their 12892 trips were
considered. The Trawl landings formed the main part of total
North Sea landings and accounted for 90 and 60 percent of total
Danish North Sea landings in weight and value respectively.
Methodology
Cluster analysis was used as a general tool for classification of
landings into groups .homogeneous with res~ect to the species
compositions. Only the relative distributions were considered.
The species composition was calculated on basis of the value of
the species landed in order to ensure that a relative small
quantity of valuable species corresponds to a non-negligible
proportion of total value. Furthermore, as the resulting
classifications may be used for analyses of effort allocation/reallocation problems, the value of the catches is a more
relevant measurethan,the quantity.
•
Fishing ground, time of landings, vessel size and other factors
are not included in the first classification, but may be
incorporated at later stage in the analysis for a final definition of the classification .
Hierarchical cluster analysis was applied, which means that the
clusters are disjoint and that one cluster may be entirely
contained within another cluster, but no other overlap between
clusters is allowed.
Input· data to the cluster analysis are vectors containing the
species composition of the landings in percent:
where
4
Landingsi
J'
-==------"-'-'=-- *1 0 0
Li
Landingsi,j
and
Xi,j
Indicates percentage in the landings of species1 for uni t j
All species landed are included in the total sum but only the
dominating species for a fleet are included in the cluster
analysis.
•
The basic idea in hierarchical cluster analysis is that each
observation (tne relative catch composition) forms a cluster by
itself. The two closest clusters are merged to form a new cluster
that replaces the two older clusters. Merging of the two closest
is repeated until only one cluster is left.
The distance between two clusters can be defined either directly
or combinatorial, that is, by an equation for updating a distance
matrix when two clusters are joined.
The direct distance between two clusters is defined as
Where
Xk
Indicatesmean vector for cluster CK
DKL
Indicates distance measure between cluster
The distance is the Euclidian distance
The combinatorial distance is
Where
5
CK
and
CL
NM
Indicates number of observation in Cluster M
For the method used the distance between two clusters is defined
as the (squared) Euclidian distance between their centroids or
means. The formula for the combinatorial distance assumes that
clusters CK and cluster CL are merged to form CM and the formula
gives the distance between the new cluster CM and any other
cluster CI'
•
Estimation of the number of clusters or main fisheries is a
delicate matter, which may depend on purpose of the analysis. For
the present analysis four objective criteria were applied for
this estimation by using plots of the number of clusters versus
pseudo F statistic, pseudo t 2 statistic, cubic clustering
criteria and the correlation (Sarle 1983, Mulligan and Cooper
1983). The number of clusters may be identified by looking for
consensus among local peaks for the cubic clustering criteria,
the pseudo F statistic and the R2 -values combined with a small
value of the pseudo t 2 statistic and a larger pseudo t 2 for the
next cluster fusion. The assumptions behind these tests may be or
are often violated. However, if all four methods indicate the
same number of clusters, this value probably is applicable and
will be selected.
principal component analysis was used as a dimension-reduction
technique for discriminating among the clusters. A scatterplot of
the cluster canonical variables visualizes the separation of the
clusters.
The SAS centroid cluster procedure was used for the calculations.
Results
•
Firstly, the 5470 species compositions of all human consumption
vessel trips in the North Sea were classified into groups defined
as main fisheries using the cluster analysis technique described
above. 9 species accounting for about 90 percent of total value
were included in the analysis.
The number of clusters were estimated using the four clustering
criteria described above.
The plots of the four criteria vs
number of clusters are shown in figure 1. All of the four
clustering criteria indicated a number of 9 clusters or main
fisheries.
The average species compositions in the resulting 9 clusters are
given in table 1. The clusters are named according to the
dominant species-(the target species) in the cluster. A cluster
with a species composition of e.g. 75 percent cod, 20 percent
haddock and 5 percent other species will be named a "Cod Haddock"_ fishery. Except for the haddock', monk and the mixed
6
main fisheries the remaining fisheries were in general rather
clean fisheries, as the average percentage of the target species
amounted to more than 60 percent of the value.
Table 1. Species composition of landings in value by main fishery of
the Danish North Sea human consumption trawlers 1988.
I
I
SPECIES
:VALUE I
ICLUSTER!
INO OFI-------------------------.--.--------.------------------------------------------1------1
I
MAIN FISHERY ITRIPsILOBSTER IPANDALUSI MONK IHADDOCK: SAITHE I PLAICE IHERRING I COD
IOTHER I % :
:---------.-.- ••••••••••••--------.------- •••• - ••••••••••••••+-- ••••••+••••• --.+.•••••••+••••••..+.••.•.••+.·····1
11
12
13
14
15
16
17
18
19
PANDALUS
PLAICE
HERRING
LOBSTER
MIXED
COD
HADDOCK
SAITHE
MONK
8511
10391
3481
1471
4861
12061
8781
2051
3101
8.31
0.6:
0.0:
86.61
10.91
0.0 I
0.11
0.61
2.81
66.41
0.01
0.01
0.51
0.01
0.01
0.01
0.41
0.21
12.51
0.4:
0.01
0.61
4.01
0.31
1.21
5.01
48.81
0.71
0.41
0.11
0.11
1.81
8.61
49.41
7.81
2.31
1.21
0.01
0.71
0.21
0.51
0.81
8.61
60.41
6.01
0.51
74.81
0.01
5.11
16.61
2.81
0.71
0.21
0.71
0.01
0.01
94.81
0.01
0.21
0.0 I
0.01
0.01
0.01
5.41
4.51
0.21
0.71
7.51
80.81
29.31
13.41
22.91
5.01 18.31
19.41 10.71
4.11 10.21
6.21 2.91
58.51 4.41
6.6: 16.71
10.7121.71
12.21 4.81
16.41 10.41
1-------------------------·------------···-------------------..--------------------------------------------------1
i_~----~~~------------------------------------------.- _ ..
..
~~~ __ i
*Note. In the calculation of the mean relative value per species, each trip has been weighted by the total value
of the trip.
The haddock, shrimps and cod fisheries were the most important
fisheries representing about 22, 18 and 17 percent of the total
value of the human consumption trawl landings in the North Sea in
1988 respectively. The haddock fishery was a combined fishery of
haddock and cod as the cod proportion was about 29 percent of the
value.
Each of the fisheries for plaice, herring and monk amounted to
about 10 percent of total value.
•
The canonical discriminant analysis is illustrated in figure 2,
where the values of the first two canonical variables, explaining
about 55 percent of total variation, are plot ted against each
other for each trip.' The figures plotted indicate the numbers
assigned to the fisheries or clusters in table 1. Cluster 3,
which was weIl separated from the other fisheries, was excluded
from the plot due to scaling problems.
Figure 2 indicates that it may be difficult to separate the cod
and haddock fisheries, and the haddock and saithe fisheries,
while the other main fisheries were weIl separated and therefore
easy to identify.
The 9 main fisheries may be described in detail with respect to
information on bycatches, when and where the fishery took place
and vessel characteristics. This is done in (Lewy and Vinther
1992) .
Two examples of the spatial distribution of the cod and haddock
. fisheries are shown in fig. 3 and 4.
7
The descriptions of the main fisheries are summarised in the
table 2.
Table 2. Features of the main fisheries of the human consumption trawl trips in
the north sea 1988.
FISHERY
VALUE IN
PERCENT
OF TOTAL
BYCATCH
VESSEL
SIZE (GT)
HOME
DISTRICT
HADDOCK
22
COD,
SAITHE
40-250
RINGK0BING,
THISTED,
LEMVIG
COD
17
HADDOCK
10-70
IserHE
5
COD,
HADDOCK
40-250
SHRIMPS
18
MONK
10-250
PLAICE
11
-
10-70
~ONK
10
COD
90-250
HERRING
10
-
100-700
SKAGEN
LOBSTER
3
-
10-500
~!ED
4
-
10-250
ALL WEST
COAST
HARBOURS
ALL
PERIOD
WHOLE
YEAR
PEAK
AUGUST
RINGK0BING,
WHOLE
LEMVIG,
YEAR
THISTED
PEAK JUNE
THISTED
APRIL/MAY
RINGK0BING
HIRTSHALS
SKAGEN
JAN. THISTED
SEPT.
PEAK
ESBJERG
HIRTSHALS
MARCH
ESBJERG
WHOLE
RINGK0BING
YEAR
PEAKS
APR. , OCT.
HIRTSHALS
JAN. TO
SEPT.
THISTED
FISHING
GROUND
THYBOR0N
TO
WEST OF
STAVANGER
EAST OF 6°
TO W.COAST
OF JUTLAND
THYBOR0N
TO
STAVANGER
NORWEGIAN
DEEP
FLADEN
EAST OF 4°
TO W.COAST
OF JUTLAND
FLADEN
SHETLAND
ISLANDS
THYBOR0N
JUNEAUGUST
TO BERGEN
CLEAVER
AUGUSTBANK
SEPTEMBER
WHOLE
YEAR
In order to obtaina manageable division of the Danish North Sea
trawl fleet with respect to metiers the annual species composition by vessel including both human consumption and industrial
landings were analyzed and grouped into sub- fleets applying
cluster analysis.
The four clustering criteria deterrnining the nurnber of clusters
indicate that there were 9 clusters. Four of these were mixed
fisheries of limited economical importance. These clusters are
further cornbined to one fishery called "other". The resulting 5
sub-fleets and their species composition are shown in table 4.
8
-
.----------
- - - - - ~- - ~ ~
Table 4. Species composition of landings in value by sub-fleet.
The Danish North Sea trawl fleet, 1988.
ISub-fleet by
Itarget species
I
INO OFI
I
I
lIND -I
ITOTAL I
% I
IVES- ILOB- (PAN- 1
IHAD- ISAIT-IPLAI-IHER- I
ISTRI-l
I % I
of I
ISELS ISTER IDALUSIMONK IDOCK I HE I CE IRING I COD IAL
IOTHERI
I valuel
..---+.-...+-..--+-----+.-.--+.----- -----·1
I-----------·-·--+·····+-·---+·-·--+--~--+·----+·-·--+
ICod, haddock
1431
1.61
4.11
I
I
I
I
I
I
I
I
I
1
421
0.51
0.31
0.31
0.31
I
I
IPlaice, ind. l.
I
I
I
1
IHerring,ind. l.
1
1
Pandalus
I
I
IIndustrial
1
1
I
I
1.51
1
I
0.8:
8.21
I
1
2.81
0.21 34.81
I
I
I
I
0.0161.91
0.01
I
1
1.71
I
1
I
I
401
0.31
0.61
0.0165.61
I
1
1
1
1
1
I
1
I
I
I
1
1
1
1
1
681 11.51 61.51
8.91
1.11
1.91
1.41
0.01
I
1
I
I
I
I
I
I
I
I
I
I
I
I
I
1
2541
0.41
0.91
1.2(
1.31
1.11
1.41
1.11
I
I
I
I
I
I
lOther
•
I
I
9.6: 20.21
I
I
911 29.31
1.01
I
I
3.11
I
I
0.21
I
I
1.81
I
I
4.41
I
I
0.91
I
1
5.71 12.81 100
I
1
I
I
5.41 13.01 18.11 100
I1
1
1
I
I
1.0 1 20.81
7.81 100
1
I
1
1
4.71
1.41
7.51 100
I
1
I
I
3.0180.51
I
I
I
I
I
I
I
I
9.31100
I
1
4.01 37.61 17.71 100
16.41
I
1
2.41
I
I
4.01
I
I
3.81
I
I
71.91
I
I
1.61
Table 4. indica~es that the vessels only targeting for industrial
purposes were the far most important one accounting for about 72
percent of total value of the trawl landings. The second most
category was the human consumption trawlers targeting mainly for
the roundfish species (about 16 percent of total value) . The rest
of the vessels participated in mixed industrial and human
consumption fisheries of plaice, herring and pandalus taking
about 12 percent of total value.
The descriptions of the sub-fleets by metiers are summarized in
the table 5.
Table 5. Features of the main trawl sub-fleets by metiers in the north sea 1988
METIERS
INDUSTRIAL
rINGS
COD,HADDOCK
OTHER
VALUE IN
SPECIES
PERCENT
OF TOTAL
72
COD,PLAICE
16
PANDALUS
4
HERRING,
INDUSTRIAL
LANDINGS
PLAICE,
INDUSTRIAL
LANDINGS
OTHER
4
!MONK, SAITHE
PLAICE
LOBSTER
IMONK
COD
OTHER
SPECIES
2
COD, OTHER
SPECIES
2
LOBSTER
IND.HERRING
OTHER SP.
VESSEL
HOME
SIZE
DISTRICT
GT
100-700 ESBJERG,
LEMVIG,
RINGK0BING
ALL
RINGK0BING
THISTED
HIRTSHALS
10- 20 SKAGEN
30- 80 THISTED
90-150 HIRTSHALS
100-500 SKAGEN
FREDERIKSH
R0NNE
10- 70 ESBJERG
HELSING0R
RINGK0BING
ALL
LEMVIG
FREDERIKSH
SKAGEN
9
PERIOD
APRIL-NOVEMBER
PEAK: MAY
ALL MONTH
FEBRUARY-JUNE
MAY-DECEMBER
MARCH-NOVEMBER
AUGUST-JANUARY
Discussion
The cluster analysis technique used for classification .is an
objective method providing results with respect to identification
of main fisheries and "metiers" which seems to be in general
agreement with our knowledge of the fishery.
The method,seems to be reliable for identification of groups for
which there are many observations and for which the target
species amount a high proportion of total catch.
•
Mixed fishery groups characterized by many species of equal value
are more difficult to identify because the resulting clusters in
these cases are sensible to variations of the species compositions and the selection of both the number of clusters and the
measure of distance between clusters used in the analysis. The
only solution to this is a detailed knowledge of the fishery,
which can serve as guideline for the definition of main fisheries
or metiers .
Management considerations requires that the fleets should be
divided into operational units, the sub-fleets by metiers. It
should be noted that the purpose of the management may affect the
definition of these units. For instance regional conditions may
require that the fleet is considered by port and more detailed
local metiers than defined above have to be identified.
The above metiers classifications of the fleet are based on the
annual species compositions of the vessels. It would an improvement if the seasonal variation of the vessels could be included
in definition of the metiers. This has been tried by considering
both monthly and quarterly species compositions of the landings.
However, the change of seasonal target fisheries for the Danish
trawlers seem to so complicated that it has not been possible to
obtain an operational fleet division.
'
•
The classification of individual trips is not such an operational
division. Instead, this classification may be considered as a
general description of existing fisheries. Besides the corresponding catch and effort data may lead to better estimates of
catchability of the target species compared to standard estimates
in the IeES area. The reason for that is that the latter
estimates usually include effort, which is not relevant for the
species in question. However, improved estimates of catchability
do not necessarily lead to better estimates of terminal fishing
mortality rate when tuning VPA' s. This is due to that fleet
catchability and the more accurate estimates may be subject to
considerable fluctuations in time, which do not result in better
estimates of terminal F for most tuning methods. Anyway, it is
worthwhile for a range of year to try estimating catchability by
target species from main fisheries identified as described above.
10
-----------------
--- - - - - - - - - - -
•
References
Anon. 1991 a. Report of the meeting of the STCF working group on
improvements of the exploitation pattern of the North Sea fish
stocks. June 1991.
Anon. 1991 b. Report of the working group on fishery units in
sub-areas VII and VIII. ICES CM.1991/Assess:24.
Biseau, A., and Gondeaux, E. 1988. Apport des methodes d'ordination en typologie des flottilles. J. Cons. int. Explor. Mer, 44:
286-296.
Laurec, A., Biseau, A. and Charuau, A. 1991. Modelling technical
interactions. ICES mar. Sci. Symp., 193:225-236.
..
Lewy, P., and Vinther, M. 1992. A bioeconomic model of the North
Sea multispecies multiple gears fishery. EC FAR project ma.1234
1992 and Danmarks Fiskeri og Havunders0gelser.
Mesnil,. B., and Shepherd,
structured model for the
and/or landing sizes
fisheries. J. Cons. int.
J. G. 1990. An hybrid age- and lengthassessment of changes of effort, mesh
in multiple-species,
multiple-fleet
Explor. Mer, 47:115-132.
Milligan, G W. and Cooper, M. C. 1983. An Examination of
Procedures of for Determining the Number of Clusters in a Data
Set, College of Administrative Science Working Paper Series 8351, Columbus: The Ohio State University.
Murawski, S. A., Lange, A. M., Sissenwine, M. P., andMayo, R. K.
1983. Definitions and analysis of mUltispecies otter-trawl
fisheries off the northeast coast of the United States. J. Cons.
int. Explor. Mer, 41:13-27.
..
Murawski, S. A., Lange, A. M., and Idoine, J. S. 1991. An
analysis of technological interactions among Gulf of Maine mixedspecies fisheries. ICES mar. Sci. Symp., 193:237-252.
Sarle, W. S. 1983. The Cubic Clustering Criterion, SAS Technical
Report A-I08, Cary, NC:SAS Institute Inc. Raleigh, North
Carolina.
11
•
C 400
P 3000
s
u
e
b
I
C
300
•
C
I
•
u
s
t 200
e
r
•••
•
0
T
T
52000
•
q
0
r
e
n
9 100
d
•
i
.
0
t
1 T
T
i
•
T
s
t
•
0
T
T
I
Number
5000
0
F
F
F
4000
F F
Clusters
r
F
r
F F F F
F
r
r
F
0.8
r
r r
F
0.7
d
,
R
0
3000
SO.6
q
u 0.5
F
t
01
F F F
0.9
F
s
1
6 8 10 12 14 16 18 20
4
Nultber
F
F F
r
T
1.0
F
p
2
Clusters
01
r
T
1
0
6 8 10 12 14 16 18 20
2 4
1
T
T T
c
-100
0
r
T
0
•
i
e
u
1
t
r
S
1
S 1000
C
n
1
u
•
i
t
e
r
1
•• • •
• • Ud
•
Q
,
r
e 0.4
~ 2000
d
I
s
t
0.3
,
~ 1000
F
0.2
F
F
F
O. 1
.
0.0
0
0
2 4 6 8 10 12 14 16 18 20
Nultber
01
Clusters
0
2
4
6
8 10 12 14 16 18 20
Nu.ber or Clusters
Figure 1. Plots of four clustering criteria V~ the number
of clust'=rs for the Danish human consumption tFips.
.
plot1 • cluster:3
15:13 Thursdav, November Z1, 1991
Plot of CANZ'"CAN1.
s~t
I1 v.tue of CLUSTER.
I
I
CANZ
12.5 •
I
I
I
I
10.0 •
I
I
I
I
7.5 •
I
I
I
I
5.0 •
I
I
I
I
2.5 •
I
I
I
I
0.0 •
I
I
I
I
11
,",
"""
,,',',111
'1111111111
1111111111
9191111111111
," 19" ,"", ,"
11111,""", 1
11 119911 111111 44
111111 11119 991 551 4
1 9 99199 19 111 4 4
1 991" 9919 995 4 4 4
1119199999991 55 94 44444
11 ;;;;;;;;9
1 5 4 4444
99999 99999
5
9
444
9 9 ;;;;;;99 99 999 549 4 5
9;;;;;;;;;;;9 9 9 99444 44 4
9;;;;;;9;;;;9 99 5 5 44944 5
99;;;;99 9999 995999454 4 9 4
99 9 9999 9
999 9555 9 5
9 999999 98 9 99 59559 9 9 99
9 99 999 ' 88 99
99559 58 8 5
9
99
8 97888 57 5 58
6 6 7 88 9 897555588
78 8
6 67 n6787898m8 5 8 55n
7 8
9
9
5
8878888n2Z2758 mss
5 5
n 7 nn888888888 78S6mn 77587 5755555
6 m8nn888mnn76n66S775775SSSSSSSSS
66677776m676n6676666SS555555S5S2222SS5S
6 6 67
·2.5 •
I
I
I
I
66666666666666666666662662222222222222
666666666662266222222222222222
2 222222222222
22
·5.0 •
_..•-.. -•....• ..•....•...-•.. _-+._..•....•.. _.+._.-•..•.•....•....•..
~_
'3.2 ·3.1 ·3.0 ·2.9 ·2.8 -2.7 '2.6,·2.5 ·2.4 ·2.3 ·2.2 -2.1 ·2.0 -1.9
CAN1
Figure 2. Principal component analysis of the Danish human consumption trawl North Sea fishery. Main fishery, indicated by the numbers 1-9, plotted for the first and
second canonical variable.
•
1
quo,t.,
Woo
2
f"".,y:Cod
8S4 tonn"
COD:
p.' .q':
catch
Quo,I.,
COD:
Ion n ••
IS8
....
Woo
eoteh
,u.......TL''T''"'T''""T'...,..CLr''""-r.u,~;=~. ,. " •• •• •• •
,ur''r''T''""T'"''TUy'""-r.u,.....,~::l:I. , . , . ' . •• •• •• ' . . t
..
•...-+++-+-i-if--I-t-tt'1--I-++++-t--t--t--il~'
fi.".ry:Cod
Ion" ••
p.r 'Q'.:
JS8
1726
I
••
.
..
..
"1-++I&-+-+-+-l-4I-W
.'t-t~;jft-l-+--+-t-+~
.
~
''I-+++-t--+-+-l-1-t11...~11
, h[Jo"+++-+-+-l-iI-f--tj
,
r
.. I
...
...••
...
..
.....
t
••t,...a--
I
;;
r
• • • • Cf
ILJ 1-\"
•.
• 0 ~,
I- ~
~Jtt'V ~ ..
o~lN,->
••
olIi
..."r+-il-.t-t-+-+-t....;..,..
~V
H--+-+--t--t-+-'-t-0+:::+:::-f'
\\
fi.ht'y:Cocl
972 tann ••
t 2 8
Ion 1'1 e
pe' I q , . :
,.u.-rUTL'Y-y......u.,.I..I.ru,oU,o:.l......
-1.' "
"
t. ,. •• 11 • t • t
W
-t-t-l-+-t-l-i-.:'::i"
\
V""\ •
f
I • . I}
..
..
..
''\-+++-+-+-4I-I-I-tt• 'I-+-+-;M--+-+-l--I-+-+f1
''t-t-H~'t-f-+-+-t-+-~
''''-+++-+-+-l-;I-t-tc'.lI
~
~,t..::.
"
''I-i:t-+-t-t-+-+-t-+-+''''ltf'1
,-lt-++-+-'+-+++-+",,\'
1A00
..
..
.
5
rl" ""
..
,
4 qua,ter
COD:
I
I
t~
o.
J quort.r
COOl
lIo. ealeh
I'
..
''!-++++-+-+-l-if--'H,..
• 't-+-++++-+-+-l-if!"~
eate"
fi.he,y:Cod
S7S ton" ••
p. r
"qr.:
t3S
,.....r'-'. . . .YJ.,...,'-yL...,.....,..""T....
~;b~'.
" ,. ,. c,
u
••
tonneo
'I C' ••
''I-+-+++-t--+-1-l~~
I
...-+-+++-t--+-i-l-I!""~
..
''''-+-f-1M-+-+-+-1-if-HO
,~.........I;;*"++-+--+-+-~
'~-+--+-+-+-+-+-+-+-+I'!~"
r'1o ~
1"'1 ..
•J..,.. ~}-t-+-+++-+-+-1-:-ir.·:+-+-r1l'rl:--\.%
,,/
.
0'
•
0
•
• • • • .1
/
•• v:
.",.L1H-4-+-+-+--+-+.~.+.-ftJ.
. . .•
:~ ~
..
...
....
..
\.
\
\..
-'
••
• • I~
'r,":,,~lf
-,;
1-,..,~ ~ ..
f'
~Pv) ..
l\ 1.,.... >
~
ttJr '
-r-.;~
I1v~~,t--f.-+_~Hf~
I .. I
..
..
..
..
.
.. ,.
.••
41
_ J.-~~
~
.
,
,. _~I(
. .I .
.~
i
.
.,
..
Figure 3. The Danish North Sea human consumption trawl cod fishery
distributed by statistical rectangle and quarter.
..•
..
•
110&
2
Fi,II.,)':HOddoCk
1102
tann.,
1 Quo,t.,
HAOOOCK:
p.'
eoteh
IQr.:
tOlln.1
204
..,...y....~~ •• " ,. •• •• • t
,.u.~,LI~YJ............
rilh"'1:Hodcloek
910
ton".,
1101
p.'
eotch
~.,
.
l~
•
'1-++~-+-+-+-=4-1-1~
:,
-•
,
.
.
a.
aa
11
qllo,t.,
..
•• ,. ' • •••• • t
r'-"-~..,.uYJY'y.y.y..y~~~
".
..
''!-+-+++++++-H,.
•'I-t-~HH--+-++-!li~"
';-+-H~+++++-I-lO
• 'I--f-+=~If-I-++-lH-~
•"t---lH--I-H-+--HH/!/,:'
: '1-_-0--+-1--+-.-+-+-+-1.I-+-d
i
.,)
••t,..L-
......
••
.....
....
..
::
• • ••
. ..,
••
..
~
/" ,L:.
t'1..../
.."
I
'"'
114-lH--I-H-:H=-~~~J Lf- \
?
t
..
..
..
• ~ •
/f~ ~
\-+-+-+++.+++.+.-+'-1 ~ \--(
'\
• • \ ~ i '3J •
""~4-+-H-f-H-+:--+.r4t, ~ ~ ••
v..
'r-'~
.J.. ,,..+-..j.....J-4fV
/;
41
~ ./~
:1=-..J...~:;
1t
.~L-j~l.o",.
..
T
..
..
7
~
..
..
,.
..
.
..
·•
~
·
• •
•
t't
.
..
..
.
a.
f7
t
•••••••
.J::t-r'
~~
I
•
QlIO,t.,
4
HAODOCK:
1101
,
~
.tI
~~l,.::
.
a. -
.. ..
.
Fi,h.,)':HOddOCk
1157
tonne.
HAOOOCK:
'
NOl cott" pe, ' Q ' . :
336 tonne.
J
.
..
..
.
••
~
·· • . r
..
.....
tann.,
141
, ••••••••.•• I'
•'I-+--+--+--+--+--+--+--+--~
.
.
'1-Hf-fl--+-++-HH,pl
..
,q'.:
••••
,,!-+-+-+-+++++-+1!:'
, '!-+--+--+-t-t-~HHIi"'7'
.
qllo,t.,
HAODOCK:
eotth
r'-"-,.LZ.,..Lly..............~u...u.,.u..,.,c=.!.,. ,. ,. ,. ... ••
..•
It
I I
•
I '1
,
,
•
.. "/
•J.;::L-
~
'il
/".c~
, •• -"./
• • • ••
...... U
A-+-+-+--t-t-t-~t-==-t-=~,
• • • ., l{,
(
.
..
..
• j
i/
.
·,•H-+-+-H-+-+-;-+I~l'
.
..
..••.
..
.....
...
....
••
Fia""'1:Hoddoek
1029
tannes
pe' 'Q'.:
183 tonnes
I..,
Iß
-~
..
~i\"
-;;
..
.
,.
o. ..
..
Figure 4. The Danish North Sea human consumption trawl haddock
fishery distributed by statistical ~ectangle and quarter.