Effect of Mesh Size/Type on Size Distribution and …square mesh gear was fished on a trial basis by TC 1-3 vessels in southwest Nova Scotia. This change in mesh configuration allowed

Gea

r Se

lect

ivity

/Tec

hnic

al I

nter

actio

ns S

ympo

sium

Northwest Atlantic

NOT TO BE CITED WITHOUT PRIOR REFERENCE TO THE AUTHOR(S)

Fisheries Organization

Serial No. N2314 NAFO SCR Doc. 93/120

SCIENTIFIC COUNCIL MEETING - SEPTEMBER 1993

Effect of Mesh Size/Type on Size Distribution and Catch Rates for 1991 Scotian Shelf Groundfish Fisheries

by

Mark A. Showell

Marine Fish Division, Bedford Institute of Oceanography P. 0. Box 1006, Dartmouth, Nova Scotia, Canada B2Y 4A2

Abstract

In recent years attempts have been made to reduce fishing pressure on Scotia Shelf cod, haddock, and pollock stocks through changes in mesh regulations. In 1991, proposed changes included increase in traditional diamond mesh and the use of square mesh. However, complications in the implementation of these regulatory changes resulted in a wide range of mesh sizes and configurations being fished concurrently through 1991. This situation presented a unique opportunity to evaluate the effectiveness of the mesh changes under actual commercial fishing operations.

Observations were made using sea-going Observers, who collected data on catch, size distribution, and mesh size/type. Data were analyzed from the following groundfish stocks: 4Vn (winter) cod, 4VsW cod, 5Z cod, 5Z haddock, and 4VWX+5 pollock. Size distribution of the catch, as well as catch rates, are compared between the various mesh size/types, and evaluated against theoretical retention curves.

Introduction_

In recent years attempts have been made to reduce pressure on traditional Scotian Shelf groundfish stocks through regulatory measures. One technique is to increase the minimum codend mesh size permitted in otter trawl fisheries, thus reducing the catch of younger year classes; for example in the mid-eighties codend mesh size regulations were changed from 120mm to 130mm (Waldron et al., 1985). This strategy produces a short-term loss in catch, but in theory increases over a longer term.

The determination of appropriate changes in trawl mesh regulations involves a number of steps. Retention curves from other areas and/or species are considered to develop theoretical changes in selection properties of the trawl. This is usually followed by a series of experiments under controlled conditions, on both research and chartered commercial vessels. It is hoped that results derived from these theoretical and experimental development steps will apply to the commercial fishery. Unfortunately, the conditions under which the fishing gear is used in commercial operations may seriously interfere with the selection characteristics of the mesh. These factors can range from subtle changes in the manner in which the mesh is attached to the trawl, to deliberate attempts to block the mesh openings. Furthermore, it is often very difficult to evaluate the results of the changes in commercial mesh regulations. Changes are usually implemented en mass at • the beginning of a fishing year, so comparisons of different gears within a year are not possible. Likewise, variations in age structure of the population from year to year due to recruitment make comparisons between years difficult.

The continued decline of cod, haddock, and pollock stocks on the Scotia Shelf since this time have sparked further changes in both gear technology and regulations. In 1990 130 and 140mm square mesh gear was fished on a trial basis by TC 1-3 vessels in southwest Nova Scotia. This change in mesh configuration allowed openings in the codend to remain unrestricted, thus allowing for the escape of small fish. Reaction by the fishermen to the square mesh gear was generally favourable during this period.

In January, 1991, mesh regulations were amended to allow 155mm diamond or 140mm square mesh codends only, for cod, haddock, and pollock fisheries. Concerns over the new regulations were expressed stridently by representative of the fishing industry as the fishing season progressed. This resulted in a Ministerial directive in June, decreasing the minimum mesh sizes to 140 diamond or 130 square. Further, fishery officers were instructed to enforce the mesh size regulations through warnings only.

This regulatory confusion resulted in an unprecedented number of gear sizes and configurations being used on the Scotian Shelf in 1991. This provided a unique opportunity to compare catch characteristics of the various gears, through information provided by the Scotia-Fundy Observer Program. The following report provides comparisons of catch rate and size distribution Scotia Fundy groundfish stocks fished under this regulatory regime.

Methods

All data were collected through the Scotia Fundy Observer Program. Deployments were made to both the inshore (<65'; TC 1-3) and the offshore (>100'; TC 5-7) fleets. Observers from Scotia Fundy Region are routinely deployed to monitor fishing operations in the Canadian groundfish fishery for the following stocks: 4Vn (winter) cod, 4VsW cod, 5Z cod, 5Z haddock, 4VSW+5 pollock. Coverage for small vessels was approximately 5% of the total fishing days, while offshore vessel coverage was almost 50%.

Estimates of catch were made visually on a set-by-set basis, and verified through volumetric calculations and observation of crew production. Length frequencies were taken from the unculled catch, with random selection, from approximately 50% of the sets.

Codend mesh size was recorded as stated by the bosun or Captain on each vessel. This stated size was confirmed by the observer, using a ruler As these measurements were not made using an approved gauge, they must be considered nominal.

For analysis, directed species was taken to be the species which made up the largest proportion (by weight) of the catch in any particular set. Catch and effort data were summarized for successful sets only, excluding sets where substantial gear damage occurred or where a codend window was employed. To minimize random effects, calculations of CPUE were restricted to mesh size/type combinations with more than 25 tons of catch in total.

Sampling for unculled length distribution of the catch was performed on approximately 25% of the fishing sets.

Numbers at length in sampled catches were adjusted for the size of the catch before pooling. For some stocks small numbers of length frequency samples precluded the comparison of length frequency distributions between gear configurations. Length frequency proportions were plotted only where the number of fish measured exceeded 1000; comparisons were made where two or more distributions were available.

Where data were sufficient, cumulative percent curves were calculated for each stock and mesh size/type combination. To facilitate comparison of different mesh types within a stock, two mesh selection parameters were calculated - length of 50% retention (1 m) and selection range (SR; length range between 25 and 75 % retention.

- 3

To compare observed length distributions with those expected from experimental studies, predicted numbers at age were taken from analytical assessments for the following stocks; 4VsW cod (Mohn & MacEachern, 1992), 5Z cod (Hunt & Buzeta, 1992) and 4VWX+5 pollock (Annand & Beanlands, 1992). Numbers at length were calculated from numbers at age using appropriate age-length keys. The theoretical proportion of fish retained by length group was taken from Halliday & White, 1989 for diamond mesh, and Halliday, 1993 for square mesh. Expected numbers at length in the catch were calculated by multiplying the calculated proportion at length in the population by the mesh retention proportion at length.

Results and Discussion

Distribution of vessel and gear combinations.

Distribution of the observed sets by stock and gear is summarized in Table 1. TC 1-3 vessels fished with the largest variety of gears, although very few sets were made directing for cod in 4Vn or 4VsW. Square mesh was employed almost exclusively by the small dragger fleet, predominantly in SW Nova (4X, 5Z). TC 5+ vessels as a rule fished with diamond mesh, about evenly split between 140 and >150mm.

Size Distribution - Commercial Fishery.

4Vn (winter) cod

This stock is fished predominantly by TC 5+ draggers, and hence sufficient data were available for diamond mesh codends only. 1,650 tons of catch were observed in total (Table 2), with over 20,000 fish measured. A good distribution over the three mesh sizes (130, 140, & 155 mm) codends used (Fig. 2).

In general, distribution of the catch by length was as expected, with 130 mm mesh showed a definite trend to select smaller fish over larger mesh (4, of 48 vs 50). However, the difference was less clear between 140 mm and 155 mm mesh (Table 3, Fig. 7).

4VsW cod

Observer coverage of this stock was principally on large TC 5+ vessels. Almost 5,000 tons of catch was observed over the course of the fishery (Table 2), from which over 40,000 fish were measured. The most common codend mesh used were 140 and 155 diamond, although 130 diamond accounted for slightly over 10% of the catch. There was no significant use of square mesh.

Selection parameters for the 130 and 140 mm diamond mesh codends were identical, with an 4, of 49 cm and selection range of 10 cm. Sets with a mesh size of 155 mm diamond showed a shift of 3 cm in 40, to 52 cm (Table 3, Fig. 8).

5Z cod

The Georges Bank fishing ground (5Z) is fished principally by small draggers (TC 1-3). In total almost 400 tons catch was observed for this stock, with this fleet accounting for 75% of the total. 140 mm diamond mesh was used most frequently, and accounted for 60% of the total catch. 130 mm and 140 mm square mesh was used to take 29 and 98 tons, respectively (Table 2).

Mesh selection trends for this stock are unusual. Square mesh, both 130 mm and 140 mm, appears to retain smaller fish than 140 mm diamond mesh. This is counter intuitive, and difficult to explain (Fig. 9). 1,o for each mcsh size/type is high, ranging from 58 cm to 67 cm (Table 3). These values are higher than those reported for cod (Halliday, 1993).

5Z haddock

Total observed landings in 1991 for this stock were the lowest of the five studied, at 123 tons. Most coverage was on small vessels (TC 1-3), where three gear types accounted for slightly over 100 tons of this catch; 140 mm diamond mesh most common, followed by 130 mm and 140 mm square (Table 2). Length sampling amounted to a total of 4,337 fish.

4 -

For this stock square mesh showed a distinct selection pattern for larger fish, with an 4 0 of 53 vs 55 for 130 mm and 140 mm square mesh. Little difference was seen in mesh selection characteristics between the two types of square mesh (Table 3, Fig. 10).

4VWX+5 pollock

Diamond mesh codends of 140 mm and 155 mm predominated in this fishery, and accounted for over 95% of the observed catch. Sampling for these two gear types was extensive, with almost 50,000 fish measured in total.

Mesh selection parameters are presented in Table 2 and Fig. 11. While the mesh type in each ogive is diamond, the 15, for each is identical at 61 cm. However, the selection range did change between types; 155 mm diamond mesh showed a narrower range at 11 cm vs 14.

Comparison with calculated length distributions.

Sequential population analyses and age length keys were available for 4VsW cod, 5Z cod, and 4VWX+5 pollock. Numbers at age were converted to proportions at length and 'fished' with each gear's selectivity pattern. Results are presented in Fig. 12 for all mesh sizes/types were a comparison was possible.

4VsW cod

Based on theoretical selection characteristics, a difference in /„ of approximately 3 and 10 cm should be seen for 140 mm and 155 mm mesh respectively, over 130mm (Fig. 12a). Calculated vs observed cumulative percent curves for this stock are shown in Fig. 13. The ogives for 130 mm mesh showed close correspondence. However, at 140 mm mesh, the l, o differed by 2 cm between calculated and observed; at 155 mm the 1, was shifted only 3 cm over the 130 mm value. Hence, although the 155 mm mesh tends to select for larger fish, the difference is much smaller than might be expected from theoretical calculations.

5Z cod

Calculated and observed distributions bear very little resemblance to each other, particularly at smaller lengths. This result may be an artifact of the distribution of the population, as almost 60% by number are below 50 cm. In addition, sample numbers are small relative to the other stocks presented. However, it is worth noting that in an analysis of similar commercial data from the 1990 fishery, a similar pattern was noted for this stock (Showell, 1990).

4VWX+5 pollock

Results of theoretical distribution calculations are shown in Fig. 12c; for this stock a mesh increase from 140 mm to 155 mm diamond mesh should produce a change in 1, o of approximately 2 cm. At 140 mm the observed curve showed close correspondence with the calculated curve (Fig. 15). This correspondence was not apparent with a mesh size of 155 mm; however, a narrower selection range was evident.

CPUE

Changing mesh size to allow the escapement of smaller fish is in turn increases the fishing effort required to realize the same catch. Under these conditions a drop in CPUE should be expected as mesh size increases. Similarly, square mesh should show a lower CPUE than diamond mesh of equivalent size. For three stocks (4Vn & 4VsW cod, 4VWX+5 pollock) this was generally the case (Fig. 16), with CPUE decreasing with increasing mesh size. For stocks where comparisons were made between diamond and square mesh types (5Z cod & haddock), the results were less conclusive (Fig. 17).

References

Armand, C. and D. Heartlands 1992. Assessment of pollock (Pollaclaus virens) in Divisions 4VWX and Subdivision 5Ze for 1991. Can.Atl.Fish.Sci.Advis.Comm. Res.Doc.92/44.

- 5

Halliday, R.G. And G.N. White 1989. The biological/technical implications of an increase in minimum trawl mesh size for groundfish fisheries in the Scotia-Fundy region. Can.Tech.Rpt.Fish.Aquat.Sci.No. 1691 153pp.

Halliday, R.G. 1993. Size selection of Atlantic cod, haddock and pollock (saithe) by otter trawls with square and diamond mesh codends of 130-155 mm mesh size. In press.

Hunt, J.J. and M.-I. Buzeta 1992. Status of the Atlantic cod stock on Georges Bank in unit areas 5Zj and 5Zm, 1978-91. Can.Atl.Fish.Sci.Advis.Comm. Res.Doc 92/48.

Mohn, R. and W.I. MacEachem 1992. Assessment of 4VsW cod in 1991. Can.Atl.Fish.Sci.Advis.Comm. Res. Doc. 92/54.

Showell, M.A. 1990. Observations on the 1990 Georges Bank cod and haddock fisheries. Can.Atl.Fish.Sci.Advis.Comm. WP90/158.

Waldron, D.E., T.D. Iles and G.V. Hurley 1985. Estimating the effects of introducing a minimum 130 mm mesh regulation on the 4X groundfish fishery. Can.Tech.Rpt.Fish.Aquat.Sci.No. 1400 473pp.

TC 1 -

It Sets Stock 130 mm Dia. 140 mm Dia. >150 mm Dia. 130 mm Sq. 140 mm Sq

4Vn Cod (Winter) 12 39 4VsW Cod 3 1

4X Cod 1 10 7 79 71

5Z Cod 6 31 • 12 78 164 .

4X Haddock 7 4 1 13 16

5Z Haddock 2 28 1 48 62

4VWX+5 Pollock 19 62 23 14 56

TC 5 - 7

ti Sets Stock 130 mm Dia. 140 mm Dia. >150 mm Dia. 140 mm Sq.

4Vn Cod (Winter) 66 176 56 3 4VsW Cod 122 781 741 5 4X Cod 13 29 5Z Cod 76 27 4TVW Haddock 68 28 4X Haddock 10 14 5Z Haddock 54 26 4VWX+5 Pollock 5 508 1,326

Table 1. Distribution of observed sets by main species caught and gear type.

- 6

Stock 130 mm Dia. 140 mm Dia. >150 mm Dia. 130 mm Sq. 140 mm Sq.

4Vn Cod CPUE (t/h) 4.96 1.78 2.00 1.07 % Disc. 4.8 8.9 3.9 34 # Sets 78 215 56 3 Catch (t) 605 818 227 8

4VsW Cod CPUE (t/h) 2.56 1.44 1.28 1.14 % Disc. 3.4 9.7 4.8 3.3 # Sets 125 781 742 6 Catch (t) 606 2265 2066 12

4X Cod CPUE (t/h) 0.31 0.35 0.16 0.14 0.12 % Disc. 0 OA IA 0.1 0.1 # Sets 1 23 36 79 71 Catch (0 .4 16 ' 13 36 26

5Z Cod CPUE (t/h) 0.28 0.77 0.22 0.10 0.19 9b Disc. 0.4 0.2 /6 0.5 0.1 # Sets 6 107 39 78 164 Catch (t) 5 231 23 29 98

4TVW Haddock CPUE (t/h) % Disc. # Sets Catch (t)

4X Haddock

0.38 3.7

66 55

0.21 5.1

28 12

CPUE (t/h) 0.78 0.35 0.27 0.13 0.17 % Disc. OA 0.5 3.0 4.9 0 # Sets 7 14 36 13 17 Catch (t) 7 14 7 6 10

5Z Haddock CPUE (t/h) 0.33 0.24 0.28 0.21 0.14 % Disc. 6.2 0.5 0 0 0 # Sets 2 82 27 48 62 Catch (t) 3 52 16 25 27

4VWX+5 Pollock CPUE (t/h) 2.2 1.26 0.88 0.43 0.46 % Disc. 0.2 0.8 0.6 0 0.7 # Sets 24 570 1349 14 56 Catch (t) 71 1489 2559 13 59

Table 2. Catch, CPUE, and discard proportion for Scotian Shelf groundfish stocks.

- 7 -

stak narah 1 ia SR 4Vn cod (win) 130D 48 12

140D 50 9 155D 50 12

4VsW cod 130D 49 10 140D 49 10 155D 52 . 11

5Z cod 140D 67 19 1305 58 12 1405 63 , 12

5Z haddock 140D 53 8 130S 55 8 140S 55 7

4VWX+5 140D 61 14 pollock 155D 61 11

3 Table 2: Length of 50% retention (I,o) and selection range (SR) for observed commercial catches of

Scotian Shelf groundfish species in 1991. Data are grouped by mesh size and type

Figure 1. NAFO statistical divisions in the Scotia-Fundy Region.

44 03 St 44 DS 70 74 75 51 55 417 55 100

7624761439 z 4050 54 513405570 74 7592311094

34 TO 45506455a

length (cm)

Figure 2: Length frequency distribution for 4Vn cod (winter); commercial fishery using 130mm diamond, 140mm diamond, and'>150mm diamond mesh OTB gear.

yi

22 NM 31 N II MC 24 N NOS 70 N 42 es NC N IN

N N N 34 N 42 44 02 N N 0 N /0 31 N OE N C 44 N

>150 die 18107 fish

•

12 N 10 Y II 47 40 SOFA 10 N /0 72 02 43 SO

length (cm)

Figure 3: Length frequency distribution for 4VsW cod; commercial fishery using 130mm diamond, 140mm diamond, and >150mm diamond mesh OTB gear.

44 411 4 w b Ot 01 /2 At Kr 0; w R n ?CC 44 Me /411>

— 10 —

32 as 40 44 .48 o Ea w e4 N 72 70 40 14 ee Se le 103 Me 10e In 110 IM

30 42 4 0 L; e2 ee 70 II 70 IP

length (cm) SC 54 14 103 710 Ill 11$

Figure 4: Length frequency distribution for 5Z cod; commercial fishery using 140mm diamond, 130 siluare, and 140mm square , mesh OTB gear.

II a a 41 • a • a • OS 72 72

9 a •

14 4 a • 0

12

length (cm)

Figure 5: Length frequency distribution for 5Z haddock; commercial fishery using 140mm diamond, 130 square, and 140mm square mesh OTB gear.

9 34 30 42 44 0 64 0 132 70 74 76 32 63 10 31

4VWX+5 pollock 140 cif a 13892 fish

17 77 If TY II 77 112 IT

0 130 die: 1(50)4413; SR=12 0140 Ma: 1(50)=513 SRA 4155 dla: 1(503441 1314.412

121

0 ° co

°

E

- 12 -

30 11 31 41 47 61 II E3 0 67 71 76 A 0 67 II 13

Figure 6: Length frequency distribution for 4VWX+5 pollock; commercial fishery using 140mm diamond, >150mm diamond, and 140mm square mesh (YTS gear.

length (ern)

Figure 7: Cumulative percent ogive for 4Vn (winter) cod; commercial fishery using 130mm • _______A:amn.A_1:4/11-ntmrliamnnti_onti_-...1A(Imm_riiamnnri-mPeh_nenterPar

4VsW cod

• 130 dla: 1(50)049; SR=10 * 140 din: I(50)=49; SRa10

155 dle: 450P252; SPIal 1

• 34 44 ID 114 03 10 04 N 10 104 101 113. 10 110

5Z cod

140 Ol4: 360107; 813.19 130 moo 1(30)=30, 911.12

o 140 op: 1(50163; SR=12

at ° 0

E

5Z haddock

• 140 Olt 1(30163; 1131.4 0130 p:1(110156; I3Roll 6 140 p:1(60156; SR.?

103

cum

ula

tive

%

- 13 -

• V 41 4? 12 V 0 07 77

length (cm)

Figure 8: Cumulative percent ogive for 4VsW cod; commercial fishery using 130mm diamond, 140mm diamond, and >150mm diamond mesh OTB gear.

length (cm)

Figure 9: Cumulative percent ogive for 5Z cod; commercial fishery using 140mm diamond, 130mm square, and 140mm square mesh OTB gear.

St 34 1P3 00 44 40 41 40 11 14 40 0 44.1170 74 70 0 10

length (cm)

O V

0

IS • E

°

Figure 10: Cumulative percent ogive for 5Z haddock; commercial fishery using 140mm diamond, 130mm square, and 140mm square mesh OTB gear.

cum

ula

tive

%

10

- 14 -

4VWX+5 pollock

s140 RR: 1150)=61: SRA* *155 di: 1150P31; SR=11

length (cm)

Figure 11: Cumulative percent ogive for 4VWX+5 pollock; commercial fishery using 140mm diamond, and 150+mm diamond mesh OTB gear.

170

4VsW cod

theoretical distribution 6 ▪ g 8 819 : 2

o A A

o a a a *

a 0

0 0 o 11

a o

a A

0

OC

0 —6,10140-41"*Loo

0 130 diamond 0 140 diamond

155 diamond

100

o

fa s'

3 40

20

5Z cod

theoretical distribution a0

° a

A0 SO 190 I O0

a

a

11

0

0

a o 0 4.

0

0

a 140 diamond o 140 square

• 0 a OS 11111111M iiiiii •..,.. iiii 1,1""1 ■ ■

40 44 03 01 SO a 70 75 SO 46 a SS 1I0 105

length (cm)

00

1CO

20

100

w

o 40

— 15 —

zs a Is a a so a to a io II to a to a no

length (cm)

100

4VVVX+5 pollock

theoretical distribution O 0

0

0

a 0

a O 0 140 diamond • a 155 diamond

0

0 A 0 0 a A A

• 75217640050E4 405610 7540 4610

length (cm)

0 A

a

a

Figure 12: Theoretical catch numbers at length for 1991 4VsW cod, 5Z cod, and 4VWX+5 pollock, calculated from population numbers froin SPA and mesh retention proportions at length.

lb.

100

I 0

a". 70

(5)

>

W

E 45

20

10

cum

ula

tive

%

no

220

03

40

2 0

4VsW cod 130mm diamond

o actual ▪ calculated

JJ I JJJJJ II J

4VsW cod - 140 diamond

tl o actual e calculated

it JJJJJJ I JJJJJJ 1111/11111

46 50 55 03

length (cm) 70 75 50 OS IC 95 10

120

4VsW cod - 155 diamond 15C

0 O

e actual calculated

E o

20

JJJJ 11111

O 56 40 OS 50 65 SO 05 70 TS CO 45 NO 55 10 21

50 25 CO U to 75 10 a BO 96 10

length (cm)

length (cm)

Figure 13: Cumulative percentage ogives, comparing calculated theoretical catch numbers against observed for 4VsW cod in 1991.

5Z cod 140mm diamond

•

•

• .actual • calculated

"4 4

5Z cod - 140mm square

•

•

•

0

e segue! • • calculated •

• •

-11

iiiiiii 1111111111111111.111 I 111111111111 llllllllllllll

120

ILO

20

IN

ICC

• 3

40

20

- 17 -

40 CO 10 • 40 0 70 70 0 16 60 II tal 106

length (cm)

46 • • CO 14 70 74 10

CO 04

ICO IC6

length (cm)

Figure 14: Cumulative percentage ogives, comparing calculated theoretical catch numbers against observed for 5Z cod in 1991.

- 18 -

110

4VWX pollock - 140mm diamond 10

A actual calculated

So

;A 10

> et,

7 e0 E • 40

413/

30

20

10

26 27 II UUn 4351 5365 PO 51 03056707 1 el S3 S6 SI • 01

cum

ula

tive

%

length (cm)

Figure 15: Cumulative percentage ogives, comparing calculated theoretical catch numbers against observed for 4VWX+5 pollock in 1991.

- 19 -

„ 130 140 150

diamond 130 140 150

diamond 140 150 diamond

Figure 16: CPUE (tons/hr) observed in 4Vn cod, 4VWS cod, and 4VWX+5 pollock directed fisheries. Data are grouped by mesh size.

a 0

140 130 140

140 130 140 diamond square

diamond square

Figure 17: CPUE (tons/hr) observed in 5Z cod and haddock directed fisheries. Data are grouped by mesh size and type.