Engineering Bycatch Reduction in West Indian Fish Traps:

STFA CRP Trap Vent Study

CRP Project No. NMF4540101

Prior Escape Vent StudiesSource Vent Height

(inches)Vent

Width (inches)

Trap Mesh SizeFish/ Trap

Haul (Control Traps)

Munro (1999) 2.76 1.10 1 inch 10.03.15 1.183.54 1.303.54 0.98

Johnson (2010) 7.87 0.98 1 inch 11.815.3/4 0.98

STFA Pilot (2008) 6 1 2 Inch 4.654 14 2

STFA CRP (2010) 6 1 3/4 2 Inch 5.445 3/4 1 1/2

5 3/4 118.0 15 3/4 1 3/8

9 3/4 1 3/8

5 3/4 1 1/4

Olsen, Dammann and Laplace (1978) 1 inch 32.8 1 by 2 inch 17.5 1.5 Inch 1.8

1 by 16 inch Edge 1 by 53/4 inch Vent 1.5 by 5 3/4 inch Vent 1 3/4 by 5 ¾ inch Vent

1 3/8 by 9 3/4 inch Vent 1 3/8 by 5 3/4 inch Vent 1 3/8 by 5 3/4 inch Vent

Escape Vents Tested

Summary of Trap HaulsDiving Studies Field Testing

Vent Type Initial Dive

Initial Haul

Trap Stock-

ing

Phase 1

Phase 2

Phase 3

Phase 4

Control 54 62 46 30 148 164 100

1 by 53/4 Vent 68 76 52 46 148 164

1 in Edge 68 76 52 43 - -

11/2 by 5 3/4 Vent 68 76 52 41 148 164

1 3/4 by 5 3/4 Vent 68 76 52 45 -

13/8 by 5 3/4 Vent - - - - 148 164 400*

13/8 by 9 3/4 Vent - - - - 148 164

Total 322 366 254 205 740 820 500

* Located top and bottom front and back of the traps.

Video Studies• 24 hour videos from 8 days (Renchen, 2010)• Only daylight hours were used in order to be able to

identify fish species and positions.*• Position of fish in trap recorded every 5 minutes.• In vent position study, 9 vents were placed in two

strings of 10 baited traps.• Vents were closed following final daily observations.• Each trap was observed for 20 minutes and all

escapes recorded.• Video was recorded for one trap each day.

Renchen reported that peak ingress/egress was around noon and occurred almost entirely during daylight hours.

Video Studies

Video Summary Results

St. Thomas St. John

Vent Choice Results

• A total of 28 escapes were observed or caught on video.

• Most escapement involved back vents (74%).• Most escapement involved upper vents

(70%).• Results were used to guide design of Phase 4

field testing.

Experimental Design (Diving Studies)1. Each trap string inspected by diving (species

counts).2. Estimate fish required for stocking.3. Haul non-experimental traps to collect fish

necessary to supplement for stocking.4. Haul experimental traps, measure (TL) all fish and

examine for fin clips.5. Stock traps with fish from experimental and non-

experimental traps that have been fin clipped.6. Traps set for 1, 3, 5, 7 days before being revisited.

Changes in Numbers During Trap Hauling(Initial Dive vs Initial Haul)

Control 1 by 6 Vent 1 in Edge 1.5 by 6 Vent 1.75 by 6 Vent0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

No Change

Gain

Loss

% o

f Tr

ap H

auls

Species Escapement During Hauling(Most Common Species)

Species Name Initial Dive Escaped During Initial Haul % Escape

BUTTERFLYFISH,BANDED 25 8 32%PORGY,SAUCEREYE 10 3 30%GRUNT,BLUE STRIPED 119 23 19%PARROTFISH,STOPLIGHT 21 4 19%BLUE TANG 189 26 14%COWFISH,HONEYCOMB 21 3 14%SCHOOL MASTER 77 9 12%COWFISH,SCRAWLED 56 6 11%DOCTORFISH 9 1 11%HIND,RED 57 5 9%GRUNT,WHITE 51 4 8%TRIGGERFISH,QUEEN 16 1 6%LOBSTER,CARIB. SPINY 38 1 3%TRUNKFISH,SMOOTH 32 1 3%ANGELFISH,GRAY 29 1 3%TRUNKFISH,SPOTTED 19 - 0%PORGY,PLUMA 14 - 0%ANGELFISH,QUEEN 10 - 0%

Total for all species 892 101 11%

0 1 3 5 70%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Control

1 by 6 Vent

1 in Edge

1.5 by 6 Vent

1.75 by 6 Vent

Set Length (Days)

% F

in C

lippe

d Fi

shRetention of Fin Clipped Fish

Retention of Fin Clipped Fish in Trap Retention of Fin Clipped Fish in Trap Hauls

Control

1 by 6 Vent

1 in Edge

1.5 by 6 Vent

1.75 by 6 Vent

Control

1 by 6 Vent

1 in Edge

1.5 by 6 Vent

1.75 by 6 Vent

Control

1 by 6 Vent

1 in Edge

1.5 by 6 Vent

1.75 by 6 Vent

Control

1 by 6 Vent

1 in Edge

1.5 by 6 Vent

1.75 by 6 Vent

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

New FishFin Clipped

Vent Type

% o

f Tot

al C

atch

1 Day Set 3 Day Set 5 Day Set 7 Day Set

Summary• Larger fish were found in in longer sets (p<.06)• 40-80% of the stocked fish left during the first day.

Highest escapement was in traps with the widest vents.

• For all vent sizes most (90-100%) of the stocked fish were gone within 7 days.

• Larger vents sizes had faster exit rates.• No species-specific tendencies were observed in exit

although the larger vent sizes exhibited significant changes in species composition at all set lengths.

Experimental Design• Standard St. Thomas Fish Traps

– Rectangular and Arrowhead Design.– Set in Strings of 10– Fishermen told to “fish them like they normally do” during field testing.

• Vents (2 per trap on the away from the funnel “front” side)– Control (no vent)– 1 by 53/4 inch vent– 1 inch vent along entire (18 inch) edge of trap– 11/4 by 53/4 inch Vent– 11/4 by 93/4 inch Vent– 13/8 by 53/4 inch Vent– 11/2 by 53/4 inch Vent– 13/4 by 53/4 inch Vent– Vents in Random order on trap string

Field Testing Series

1. Phase 1 field testing with vents from diving studies.2. Phase 2 Field testing.

1 by 53/4, 1 3/8 by 53/4, 1 3/8 by 93/4, 11/2 by 53/4 and control traps.

3. Phase 3 testing. 1 by 53/4, 1 3/8 by 53/4, 1 1/4 by 53/4

(top of trap), 1 1/4 by 53/4 (bottom of trap) and control traps.

4. Phase 4 testing. 1 3/8 by 53/4 placed on the front and back top and bottom to test for vent location.

Study Catch Characterization (94 species)

Category # Caught %Commercial Catch 8,562 62.7%

Bycatch 5,099 37.3%

Ciguatera 600 11.8%

Non Commercial 146 2.8%

Regulatory 23 0.5%

Too Small 4,330 84.9%

Boxfish 2,353 54.3%

Not Boxfish 1,977 45.7%

Field Testing Summary Results: General– Average Catch/Trap Haul does not stabilize until > 40 trap

hauls.– No difference between rectangular and arrowhead traps– Catch/trap haul not correlated with set length.– Catch/trap haul not correlated with temperature.– Lunar affects (catches greater and fewer empty traps

around full moon).– Seasonal peaks for most species.– Lionfish from 0 to 19th most common species at end of

study.– Mortality rate (0.031fish/day ) similar to diving studies

and Renchen (2010) study (0.025fish/day ).

Cumulative Average # Fish/Trap Haul (Control Traps)

1 10 19 28 37 46 55 64 73 82 91 1001091181271361451541631721811901992082170

2

4

6

8

10

12

14

Pilot Study

Phase 1

Phase 2

Phase 3

Phase 4

# Trap Hauls

# Fi

sh/T

rap

Haul

Species Seasonality of CPUE

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec0

1

2

3

4

5

6

7

Queen AngelYellowtail SnapperSaucereye PorgySchool master SnapperRed Tail ParrotfishStop Light ParrotfishGray AngelBlue Striped GruntHoneycom TrunkfishDoctor FishRed HindLongspine SquirrelfishSmooth TrunkfishScrawled TrunkfishBlue TangWhite GruntQueen Triggerfish

Fish

/Tra

p Ha

ul

Gear Selectivity• Where:

S is Selectivity and and are calculated iteratively via nonlinear regression. From the non-linear model (Proc NLIN in SAS) with known values of size class and percent of the sampled population for that size class. The output was estimates of both parameters, which were then used within the selectivity equation to estimate size selectivity for each size class and particularly the size class at which selectivity peaked.

Selectivity Analysis: Fish TrapsYellowtail snapper (Ocyurus chrysurus) [Traps]

Size Class

220 230 240 250 260 270 280 290 300

Sel

ectiv

ity

0.0

0.2

0.4

0.6

0.8

1.0

1.2

ObservedPredicted

Summary of Fish Life History Results

Sample Size Species Measure-

ment

L first reproduction

Female

L first reproduction Male (mm)

Min Catch (mm)

MidPoint Overlap Min

Catch and Max By catch (mm)

% Above L first repro-

duction

Average Size

Retained as Catch

Size at 50% Recruit-ment to Fishing Gear

Size at Full Recruit-

ment

57 Pomacanthus paru TL 226 268 192 202 37% 271.2 164.5 250-280

417 Pomacanthus arcuatus TL 220 240 184 193 62% 256.5 159.4 200

135 Holacanthus ciliaris TL 170 242 266.6 209.4 260

2120 Acanthurus coeruleus FL 130 110 173 221 100% 244.1 159.5 200

262 Chaetodon striatus TL 133 100% 120.4 130

168 Epinephelus fulvus TL 160 220 281.4 265.6 280

674 Lactophrys poligonius TL 165 223 267.0 185.3 220

1356 Lactophrys quadricornus TL 170 160 162 211 100% 258.0 184.0 210

843 Acanthurus chirurgus FL 170 140 210 213 100% 244.6 208.6 240

535 Haemulon sciurus FL 170-200 140 210 100% 280.9 250.7 290

27 Haemulon carbonarium FL 224 264.9

28 Haemulon flavolineatum FL 160 204 100% 267.4

1020 Haemulon plumieri FL 130-150 153 216 100% 274.5 244.9.5 270

Sample Size Species Measure-

ment

L first reproduction

Female

L first reproduction Male (mm)

Min Catch (mm)

MidPoint Overlap Min Catch and Max By catch

(mm)

% Above L first repro-

duction

Average Size

Retained as Catch

Size at 50% Recruit-ment to

Fishing Gear

Size at Full Recruit-

ment

1078 Epinephelus guttatus TL 250 214 221 94% 321.4 279.4 300

110 Pterois volitans TL 256.1 310

296 Sparisoma chrysopterum FL 240 303.5 278.2 310

414 Sparisoma viride FL 180 160-200 222 100% 312.9 259.9 320564 Calamus calamus FL 200 266.7 217.3 250513 Lutjanus apodus FL 250 225 272.6 290

25 Lutjanus buccanella FL 230-260 250-270 252 56 28% 284.6 276.5

204 Lutjanus synagris FL 180-235 150-214 248 100% 290.4 278.1 27054 Lutjanus analis FL 410 380 300 58% 411.5 259.3

Ocyurus chrysurus FL 224 188

156 Traps FL 230 100% 307.5 271.8 2901973 Handline FL 210 92% 329.8 278.1 300182 Seine FL 200 96% 320.1 270

1337 Panulirus argus CL 89 81.7 100885 Holocentrus rufus FL 135 208 100% 281.0 273 230

1610 Balistes vetula TL 235 265 223 236 96% 384.3 218.5 280

953 Lactrophry triqueter TL 258 131.6 180

264 Lactrophrys bicaudalis TL 200 219 275.6 150.9 200

Summary of Fish Life History Results

Total Mortality Rate

The annual total mortality rate (shown in equation 5

Size Frequency distributions for the most common species were analyzed for total mortality Z (equation 1).

(1)

Where: is the number of fish at time (t) and Z is the rate of total mortality.

An age specific instantaneous rate of Z was calculated by solving the Von Bertalanffy growth equations (equation 3) for t (median age of the size class in the frequency distribution.

(2)

Where: is the length (in mm) at age (t). is the asymptotic length at which growth is zero. t0 is included to adjust the equation for the initial size of the organism and is defined as age at which the organisms would have had zero size.

Solving equation (2) for t (age) allows for calculation of the age (in years) of each size class.

(3)

Species Mortality Values

Species #Lmax TL

(in mm)

Information from FISHBASE Current Analysis

kLocation Total Z

Epinephelus cruentatus 168 388 340 0.23 -5.3 USVI 1.025Lutjanus synagris 204 385 450 0.23 Puerto Rico 0.956Sparisoma chrysopterum 296 385 418 0.78 BVI 1.988Sparisoma viridae 420 402 290 0.96 USVI 0.567Lutjanus apodus 563 605 349 0.35 USVI 0.351Haemulon Sciurus 315 337 371 0.3 Puerto Rico 0.543Acanthurus chirurgus 843 318 332 0.13 USVI 0.517Haemulon plumieri 879 388 420 0.26 Puerto Rico 2.727

Epinephelus guttatus 1,356 485 568 0.12 St. John 1.037

Balistes vetula 1,483 649 600 0.23 USVI 0.966

Acanthurus coerulus 2,116 414 369 0.11 USVI 0.374

Ocyurus chrusurus (Handline) 1,874 630 502 0.139 -0.96 Puerto Rico 0.120

Ocyurus chrusurus (Traps) 230 630 502 0.139 -0.96 Puerto Rico 0.093

Panulirus argus 1,337 180 152 0.432 -.11 St. John 0.472

Field Testing Summary Results: Vents• Box fish.

Only 2 by 4 and 13/4 by 53/4 vents let out boxfish and these also had very low retention of commercial catch. Study then concentrated on release of thin bodied not-box fish (TBNBF).

• Vent height.– 1 inch by 18 had significant release of TBNBF and retention of

commercial species.– 1 by 4 released significantly less TBNBF than 1 by 6 in pilot study.– 13/8 by 9 3/4 Vent did not release more TBNBF than 13/8 by 5 3/4 .

• Vent Width. 11/4 by 5 3/4 and 13/8 by 5 3/4 both has good release of TBNBF and retention of catch but 13/8 by 5 3/4 had better release of bycatch.

• Vent Location. Top and bottom not significant but vents in front (away from funnel) side had significantly higher release rates. (Contrasts with diving observations).

1.75 by 6 Vent

1.5 by 6 Vent

1 by 6 Vent

Control (Phase 2)

1 by 41 inch Vent

Control (Phase 3)

Control (Pilot)

2 by 4Control (Phase 4)

1 by 6Control (Phase 1)

1 by 6 Vent

1.375 by 10 Vent

1.25 Vent

1.375 Rear Top

1.375 Rear Bottom

1.25 Top

1.375 Front Top

1.375 by 6 Vent

1.375 Front Bottom

1.375 Vent

1 in Edge

1.5 by 6 Vent

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

0

0.2

0.4

0.6

0.8

1

1.2

1.4Catch Too Small By Catch

Aver

age

Catc

h (#

Fish

/Tra

p Ha

ul)

Aver

age

Too

Smal

l By

catc

h (#

Fish

/Tra

p Ha

ul)

Comparison of Catch and Too Small (Not Boxfish) By catch by Vent Type

Impact of Selected Vents Control 13/8 Vents Comparison

Species Count Avg. TL Count Avg. TL Control CPUE

Vent CPUE Reduction

% Reduc-

tion Control CPUE

% Control

TL

Acanthurus coeruleus 311 192.77 75 200.09 0.755 0.182 0.573 81% 103.8%Acanthurus chirurgus 135 232.47 92 239.15 0.328 0.223 0.104 45% 102.9%

Balistes vetula 295 368.42 258 380.39 0.716 0.626 0.090 30% 103.3%Calamus calamus 62 266.82 25 279.08 0.150 0.061 0.090 68% 104.6%

Pomacanthus arcuatus 43 225.26 33 262.91 0.104 0.080 0.024 38% 116.7%Holacanthus ciliaris 15 234.27 2 205.00 0.036 0.005 0.032 89% 87.5%Holacanthus tricolor 22 185.50 8 186.29 0.053 0.019 0.034 71% 100.4%Chaetodon striatus 27 124.33 9 125.78 0.066 0.022 0.044 73% 101.2%

Sparisoma chrysopterum 55 304.36 32 307.63 0.133 0.078 0.056 53% 101.1%Sparisoma viride 55 308.89 55 324.64 0.133 0.133 0.000 20% 105.1%Lutjanus apodus 59 319.00 39 321.64 0.143 0.095 0.049 47% 100.8%

Holocentrus rufus 135 283.16 135 285.23 0.328 0.328 0.000 20% 100.7%Lutjanus synagris 19 285.95 17 287.53 0.046 0.041 0.005 28% 100.6%

Lutjanus buccanella 17 283.71 3 311.33 0.041 0.007 0.034 86% 109.7%Ocyurus chrysurus 13 347.69 28 364.39 0.032 0.068 -0.036 -73% 104.8%

Haemulon plumieri 109 278.40 106 280.00 0.265 0.257 0.007 22% 100.6%Haemulon sciurus 55 285.18 69 284.10 0.133 0.167 -0.034 -1% 99.6%Epinephelus fulvus 25 284.56 32 278.25 0.061 0.078 -0.017 -3% 97.8%

Epinephelus guttatus 148 321.09 169 324.96 0.359 0.410 -0.051 8% 101.2%Pterois volitans 22 267.82 25 282.36 0.053 0.061 -0.007 9% 105.4%

Haemulon melanurum 17 271.00 4 259.25 0.041 0.010 0.032 81% 95.7%Lactophrys quadricornis 197 224.89 273 217.19 0.478 0.663 -0.184 -12% 96.6%

Lactophrys triqueter 126 168.40 184 168.82 0.306 0.447 -0.141 -18% 100.3%Lactophrys bicaudalis 24 212.79 31 207.39 0.058 0.075 -0.017 -4% 97.5%Lactophrys poligonius 74 237.45 127 232.70 0.180 0.308 -0.129 -38% 98.0%

Impacts of 13/8 Vents on Selected Species

SpeciesL first repro-

duction Female

L first repro-

duction Male

Size at 50%

Recruit-ment to

Gear

Min Catch

% Above L first repro-

duction

% Reduct-ion from Control CPUE

% Increase

in TL

Acanthurus coeruleus 130 110 159.5 173 100% 81% 4%

Acanthurus chirurgus 170 140 208.6 210 100% 45% 3%

Sparisoma chrysopterum 278.2 240 53% 1%

Sparisoma viride 180 160-200 259.9 222 100% 20% 5%

Pomacanthus arcuatus 220 240 159.4 184 62% 38% 17%

Holacanthus ciliaris 209.4 170 89% 1%

Chaetodon striatus 120.4 133 73% 1%

Calamus calamus 217.3 200 68% 5%

Balistes vetula 235 265 218.5.4 223 96% 30% 3%

Implementation• Fishermen from Phase 4 of study are leaving

vents in. (Julian Magras, Daryl Bryan, Tony Blanchard and Danny Berry)

• Proposal to Bycatch Reduction Program.– Fishermen install vents in one string.– Port Sample vent string and one regular string.– If convinced project to supply vents for remainder

of traps.• Council/Territorial Regulation.