................
I,
ICES CM 1986/B:12,
Fish Capture Committee
Ref: SheHfish Committee
International"Council for the
Exploration of the Sea
Sess V: Gear Selectivity
SQUARE AND DIAMOND MESH TRAWL CODEND SELECTION
TRIALS ON NEPHROPS NORVEGICUS (L)
by
J H B Robertson, D C Emslie,
K A Ballantyne and C J Chapman
•
DAFS Marine Laboratory ,
PO Box 101
Victoria Road
Aberdeen Scotland
INTRODUCTION
The Marine Laboratory has been conducting selection experiments using square mesh
in the codends of trawl n,ets. Square mesh can release more juvenile roundfish than
traditional diamond mesh (Robertson, 1983; 1984; Robertson and Stewart, 1986). The
effect of square mesh 'on other species inc1uding Nephrops norvegicus is being
investigated. This report describes initial results on the selection characteristics of
65 mm diamond and 57 mm and 67.6 mm square mesh codends for Nephrops
norvegicus. The covered codend technique was used.
MATERIALS AND'METHODS
The research vessel "Aora" owned by ~arine Biological Station, Millport, conducted
nine tows.during November 1984 and 14 tows in February 1985 in the Clyde, mainly to
the north, west and south of the island of Cumbrae on commercial fishing grounds.
The test codends were attached to a typical commercial dual purpose trawl of 2,5~8 m
headline length with a upper wing mesh size of 12,0 mm, lower wing 60.7 mm, beHy
61 mm and lower extension mesh size of 64 mm, as measured with an ICES gauge of
4 kg tension.
The trawl rig consisted of 1.8 m long. vee otterboards with a single 3 m backstrop
leading into double 36.6 m wire bridles. Rubber discs of 50 mm diameter were reeved
hard packed onto the last 18.3 m of the lower bridles. In addition there was a 1.8 m
chain extension between the lower bridle and .wing end.
1
Comparisons\vere made between a 65 mm diamond and 67.5 ~m square mesh cod:~d
during November and 65. mm diamon~, 67.5 mm and57 mms,quaz:e me~h .codends durmg
February 1985. Codend length was 6.1 m with 120 meshes on the Jom round of the
65 olm diamond cases and 120 bars in the 67.6 mm and 57 mm square cases. Netting
material was single twisted polyethylene of 400 ni/kg rmmage.
Nephrops population estimates were made using a small mesh .codend cover (2.5~6 ~m
mesh twisted polyethylene twine 580 m/kg) over the codend to capture the escapmg
animaIs. The cover length was 9.2 m and thecodend was hanging freely inside with no
internal attachments to the cover. Both codend and cover were attached to the trawl
at the same join round~ The cover netthlg waS approximately one and a half tim es the
width of the test codend to minimise masking as recommended by Pope et al., (1975)
and Stewart and Robertson (1985).
Regular mesh measurements were taken on each codend and the cover with an lCES
gauge. Between 150 and 200 wet meshes were measured randomly along the codend in
the top and lower. panel imm.~d!ately after a h,,:ul. All the. codends.were irieas~red
twice throughout the experiments. The average measurements thus derlved are quoted
in the report.
All of the Nephrops from the codend and cover
and carapace length is recorded in the report~
\v~re measured at the end of each haul
•
Data Analysis Procedure
An appropriate method of analysis. for data from codend mesh selection experiments is
the fitting of aselection curve. The logistic curve is used in this analysis using the
logit transformation on the proportions of fish retained at each lengthto fit a straight
line by normal linear regression methods, and obtciizi estimates of the parameters of
the curve~ An iterative procedure using, the expected logits,
suitable \veights;
gives corrections to' these parameters; Ieading to maxhnum likelihood estiInates of the
parameters of the curve. When these parameters areestablished 50% points, selection
ranges and selection factors can bei calculated; which are then used as the basis for
comparison of mesh selection.
and
Selection curves were etrawn for each haul. Eight valid hauls were obtained in
November and 13 in February. Two hauls (one [rem] November and one from February)
were not used in the analysis due to possible severe masking of the coderid meshes .by
the cover and/o: deficient numbers of animals; gear or codend cover damage.
..
RESULTS
The dat~ cOlIected durfng November and February have been presented in separate
selection curves mainly becatise .on all the November hauls largequantities of
. Laminaria weed were collected. Internal maskirig of the codend meshes by the weed.
may·have occurred (see Discussion).
.
,.,'
,
.:,
The selection curves for each codend are contained in Figures. 1 and. 2. for the
November cruise and 3; 4 and5 fcr the February crUises. Combined Iength/frequency
distri.butions of animals retained byeach coderid for the November and February
cruises may be found in Figures 6 and 7 respectively. The valid hauls for each coderid
were combined to give asingle selection curve for each mesh size (Fig. 8 for
November and Fig. 9 for February)~
'...,
,
'
"
_
A summary of the selection features [or each codend are as follows:
No of
combined
hauls
50%
length
(mm)
Selection
Factor
Range
(mm)
Percentage
Nephrops
retained
in codend
Nov
4
17.1
2.64
17.1
84
67.6
Nov
4
39.3
5.82
19.0
40
Diamond
65
Feb
3
20.2
3.11
16.8
77
Square
67.6
Feb
4
26.3
3.88
50.6
54
Feb
6
29.7
5.21 .
12.1
59
Mesh
size
(rom)
Month
Diamond
65
Square
Mesh
type
•
Square
.57
DISCUSSION
The Feburary 50% retention length of 20.2 mm far the 65 mm diamond mesh compares
weIl with the figure of 21.1 mm for 65 mm mesh determined by Bennett (1984) and
quoted in Briggs (1984). Hillis and Earley (1982) give a 50% figure of 22 mm for 70
mm mesh and again this falls reasonably close to the February data presented in this
report.
If internal masking of the November codends did occur because of the catches of
\veed, then one \vould expect the 50% length for the November data to be lower than
the February data which was weed free. This is indeed the case for the 65 mm
diamond mesh but not for the 67.6 mm square mesh. This anomaly may be related to
the fact that masking of the February 67.6 mm square mesh codend by the small mesh
cover is thought to have occurred when quantities of shells were caught particularly in
hauls 4, 5 and 6. The heavy codend may have sagged onto the small mesh cover, thus
blocking the codend meshes for escape. It is therefore impossible to quantify the
possible unpredictable effects that the catch of weed may have had on the November
data.
It is feit that a truer measure of selection could be accomplished without covers by a
direct alternate haul comparison of Nephrops catches from the codends or better still
by comparing the two codends on a twin trawl.
Although a minimum number of hauls were achieved for all the codends it would seem
that the 57· mm square mesh codend has better selection characteristics than the other
diamond arid square mesh codends .used in' this study. The selection ranges for the
February 57 mm square and 65 mm diamond are 12.1 mm and 16.8 mm respectively.
The 57 mm square mesh has a steeper selection curve giving sharper selection than
both the 65 mm diamond and 67.6 mm square mesh. This is illustrated clearly in
Figure 9. However, more data is required to verify this result.
3
..----------~---~--------~~--
Bennett (985) has outlined the mangement problem in the Irish Sea Nephrops fishery
where the permitted Nephrops mesh si~e is .10 to 15 mm less than that allowed for
finfish~
He reviewed recent work done by Main and Sangster (1982) on separator
trawls which optimise Nephr.ops exploi~ation, while minimising finfish. by-catches.
Bennett postulates the possible use of separator trawls to solve some of the problems
in the Irish Sea. Square mesh of 57 mm seems to have better selec~ion characteristics
than the present 70 mm diamond mesh on Nephrops. A separator trawl with square
mesh in the lower codend to optimise retention of legal sizes.of Nephrops whilst
minimising discards and in the top codend a square or diamond mesh size that would
retam a marketable fish by-catch, may give a better solution to the problem than by
using diamond mesh codends with separator trawls.
Alternatively a horizontally divided square mesh codend could be used by itself. This
would be less expensive to commercial fishermen "than a ftin length separator panel
~ that can require one third extra netting than the whole tra\vl. Such ci codend could
have square mesh in its lower part .for Nephrops and a larger mesh size in the top part
appropriate for the selection. of marketable whitirig. and haddock~ Various partly
successful attempts to separate whiting and haddock from Nephrops iIi codends have
been made, Symonds and Simpson (1971), Ferris (1970), Hiliis U983; 1984)~ The front
aperture of square mesh codends can be. designed to be bigger than diamond mesh
codends, Robertson (1986). There could therefore be more room for separation which
may improve the results obtained by these workers. However, cod is a parÜcular
problem where separation is less predse.
ACKNOWLEDGEMENTS
We thank Mick Parker and the crew of "Aora" ror the exceIlent job they
the 'gear and Jim Atkinson for organisational aide
did
i~ handling
.
REFERENCES
Ferris, G.D. (970). Interim report on double storey codend Nephrops experiment. N
Ireland, MS. .
Hillis, J.P. and Early, J.J. (1982). Selectivity in· the Nephrops trawl.
1982/B:19, Fish Capture Committee.
ICES CM
Hillis, J.P. (1983). EXperiments with a double codend Nephrops tra\vl.
1983/B:29, Fish Capture Committee.
ICES CM
4
•
HilUs, J.P. (1984). Further experiments with a double codend Nephrops trawl. ICES
CM 1984/K:36, Shellfish Committee.
Main, J. and Sangster, G.I.
(1982). A study of separating fish 'from Nephrops
norvegicus L. in a bottom trawl. Scottish Fisheries Research report number 24.
Pope, J .A., Margetts, A.R., Hamley, J .M. and Akyuz, E.F (197 5). Manual of methods
for fish stock assessment. Part III selectivity of fishing gear. F AO Fisheries
Technical Paper, No. 41 Revision 1.
Robertson, J.H.B. (1983). Square mesh codend selectivity experiments on whiting
(Merlangius merlangus (L) and haddock (Melanogrammus aeglefinus (L)). ICES
CM 1983/B:25, Fish Capture Committee.
Robertson, J.H.B.
(1984).
Square mesh codend selectivity for haddock
(Melanogrammus aeglefinus (L)) and whiting (Merlangius merlangus (L)) in a
Scottish seine net. ICES CM 1984/B:30, Fish Capture Committee.
•
Robertson, J.H.B. (1986). The design and construction of square mesh codends.
Scottish Fisheries Information Pamphlet, Number 12 (in press).
Robertson, J.H.B. and Stewart, P.A.M. (1986). An analysis of leIigth selection data
from comparative fishing experiments on haddock and whiting with square and
diamond mesh codends. Scottish Fisheries Working Paper No 9, 1986.
Stewart, P.A.M. and Robertson, J .H.B. (1985).
Fisheries Research Report Number 32.
Small mesh codend covers.
Scottish
Symonds, D.J. and Simpson, A.C. (1971). Preliminary report on a specially designed
Nephrops trawl for releasing undersized roundfish. ICES CM 1971/B:6, Gear and
Behaviour Committee.
5
NEPHROPS November 1984
Hauls 1 7 8 9
65mm diamond mesh
J
J
J
A084/1 Diamond 65.0mm
A084/7 Diamond 65.0mm
1.0
1.0
0.8
0.8
0.6
0.6
0.4
0.4
0
0- 0.2
0
L
0.2
D
Q)
x x
.-(tlC
+J
Q)
Cl::
C
0
..,J
L
•
xx x
a..
0.0
10
20
30
40
50
60
0.0
10
A084/8 Diamond 65.0mm
40
50
60
50
60
1.0
x
Q)
30
A084/9 Diamond 65.0mm
1.0
D
20
x
~
0.8
0.8
C
'-
(tl
..,J
Q)
0.6
0.6
x
Cl::
C
0.4
0.4
0
0- 0.2
0
L
0.2
0
+J
L
e
a..
0.0
10
20
30
40
50
Length (mm)
60
0.0
10
20
30
40
Length (mm)
FIG 1
NEPHROPS November 1984
Hauls 2)4)5)6
67.6mm square mesh
A084/2 Square 67.6mm
"0
Q)
A084/4 Square 67.6mm
1.0
1.0
0.8
0.8
c
.-
x
(U
..>
Q)
0.6
0.6
a::
C
0
OA
L
x
0.2
0.0
10
30
20
40
50
60
0.0
A084/5 Square 67.6mm
Q)
x
x
~
0-
TI
x
x
~
x
0.2
0
•
>«
x
L
0
-
x xx
0.4
..>
Cl.
x x
x
10
20
30
40
1.0
0.8
0.8
0.6
)(
0.4
a.
0
/
)(
x
0.4
.+J
L
0
60
x
0.6
~
C
0
50
x
x
.-(U
Q)
60
A084/6 Square 67.6mm
1.0
C
..>
50
x
x
)(
0.2
)(
x
x
0.2
x
L
CL
)(
0.0
10
20
30
40
50
Length (mm)
60
0.0
10
20
30
40
Length (mm)
FIG 2
NEPHROPS Februar~ 1985
Hauls 1 2 14
65mm diamond mesh
J
J
A085/1 DIAMOND 65.0mm
A085/2 DIAMOND 65.0mm
1.0
1.0
x
x
D
Q)
0.8
0.8
C
.CO
.,J
Q)
0.6
x
c:::
C
0
0.6
x
0.4
0.4
L
0
Cl.. 0.2
0
L
0-
0.2
.,J
•
0.0
10
20
30
50
0.0
10
20
30
40
50
Length Cmm)
A085/14 DIAMOND 65.0mm
1.0
1J
Q)
)(
;/
0.8
)()(
)(
x
C
'CO
~
Q)
c:::
c
Q
-
x
)(
0.6
)(
)(
0.4
)(
.,J
L
0
Cl.. 0.2
0
e
L
0-
0.0
10
20
30
40
50
Length (mm)
FIG 3
NEPHROPS Februar~ 1985
Hauls 3 4 5 6
67.6mm square mesh
J
J
J
A085/3 SQUARE 67.6mm
A085/4 SQUARE 67.6mm
1.0
1J
0)
1.0
x
0.8
x
.-(0C
+J
0)
x
0.8
x
X
x
xX)(y'X M X
X
x
0.6
x
x
x x x
0.6
~
C
0
x
x
x
0.4
x
0.4
+J
L
x x
0
0- 0.2
0
0.2
L
a..
0.0
10
20
40
30
50
0.0
A085/5 SQUARE 67.6mm
0.8
30
40
50
1.0
x
Q)
20
A085/6 SQUARE 67.6mm
1.0
"Cl
10
0.8
xx
C
.-(0
+J
(])
~
C
0
+J
x
0.6
x
x
0.6
x
x
x
0.4
0.4
x
xxx
L
0
0- 0.2
0
II
>t
XX
Rst
X
-
x x
x
x
0.2
L
x
a..
0.0
10
20
30
40
Length (mm)
50
0.0
10
20
30
40
50
Length (mm)
FIG 4
NEPHROPS Februar~ 1985
Hauls 7/8/9/10 11/12
57mm :square mesh
1
A08517 SQUARE 57.0mm
A085/8 SQUARE 57.0mm
1.0
1.0
)(
)(
1J
Q)
0.8
0.8
0.6
0.6
0.4
0.4
c
.rn
+J
Q)
a:::
c
0
..,J
x
L
0
Q.
•
0.2
0.2
0
L
0-
0.0
10
30
20
40
50
0.0
10
A085/9 SQUARE 57.0mm
Q)
30
40
50
40
50
A085/10 SQUARE 57.0mm
1.0
1J
20
1.0
0.8
0.8
x
x
C
.rn
..,J
Q)
0.6
0.6
Cl::
C
0
x
0.4
0.4
•..,J
L
0
Q.
0.2
0.2
0
e
xX
x
L
0-
0.0
10
20
40
30
50
0.0
A085/11 SQUARE 57.0mm
"0
Q)
rn
Q)
1.0
0.8
0.8
x
X
c
/
0.4
Q.
x
0.6
x)(
L
0
x )(
x
0.6
a:::
.-0
..,J
30
20
A085/12 SQUARE 57.0mm
1.0
C
...,J
10
0.4
)(
x x
)(
x
0.2
0.2
0
)(
x
L
Cl..
0.0
10
20
30
40
Length Cmm)
0.0
50
10
20
30
40
50
Length- Cmm)
FIG 5
e--
Hau I s 2+4+5+6
O-Hauls 1+7+8+9
CFILNLO.1)
CFILNLO.2)
NEPHROPS November 1984 combined hau/s
240
180
::J)
u
cQ)
:J
(J
k65mm diamond
120
Q)
L
l.L
60
«67 ·6mm square
""Tl
1-1
24
32
40
C)
Length (m m)
48
56
.
~HaUIS
7+8+9+10+11+12 CFILNLO.1)
O-Hauls 3+4+5+6 CFILNLO.2)
O-Hauls 1+2+14 CFILNLO.3)
NEPHROPS February 1985 combined hau/s
2000
1500
~57mm square
:n
u
c
~
rr
1000
Q)
L
I.L
/
67·6 mm square
500
o-l-----1~#;2~-l-------l.-------L------L----=~==#==Q...J
27.5
17.5
22.5
32.5
37.5
42.5
"Cl
t-i
Length
(mm)
J
•
NEPHROPS November 1984
Combined hauls
Square and diamond mesh
1..0
...... ........ .....................
o
0.8
a
a
a
TI
Q)
a a
a
D
C
a
'-
(0
-+->
0.6
Q)
Cl::
C
0
a
0.4
-+->
L
0
0...
0 0.2
L
CL
.... .... ..
'
..
'
0.0
1.0
20
30
40
Length
Hau I
C.E.MeSh
Size (mm)
Length
(mm)
50~
50
60
(mm)
Selection
Total No.
Factor Ranoe mm Cod End Cover
No. in S.R.
Cod End Cover
CI
67.60
39.3:t 0.82
5.82
19.0
286
435
249
334
0
65.00
17.1:!: 1.56
2.64
17.1
1553
286
138
65
o Hauls 2+4+5+6
67.6mm square mesh
o Hauls 1+7+8+9
65mm diamond mesh
FIG 8
NEPHROPS Februar~ 1985
Combined hau/s
Square and diamond mesh
1..0
. . .' .
'
0.8
LJ
0)
c
co
+J
0.6
.0)
'0::
C
o
0.4
"
.'. .
+J
L
o
0..
o
0.2
L
0...
.' .
0.0 -+--------r-------,-------,----------,
20
1.0
30
50
40
Length (mm)
C.E.Mesh 50% Length
(mm)
Size (mm)
Haul
Selection
Total No.
Factor Ranee' mm Cod End Cover
No. in S.R.
Cod End Cover
0
57.00
29.7 ± 0.15
5.21
12.1
4785
3296
3833
2999
x
67.60
26.3 ± 0.52
3.88
50.6
9250
7800
9244
7795
<>
65.00
20.2 ± 0.55
3.11
16.8
7448
2244
1936
944
o Hauls 7+8+9+10+11+12
57mm square mesh
x Hauls 3+4+5+6
67.6mm square mesh
<> Hau Is 1+2+14
65mm diamond mesh
FIG 9