The Effects of Fish Trap Mesh Size on Reef Fish Catch off

by JA BOHNSACK · Cited by 48 — ABSTRACT-Catch and mesh selectivity of wire-meshed fish traps were tested for eleven different mesh sizes ranging from 13. X 13 mm (0.5 x 0.5″) to 76 x 152

277 KB – 11 Pages

PAGE – 1 ============
The Effects of Fish Trap Mesh on Reef Fish Catch off Southeastern JAMES A. BOHNSACK, DAVID L. SUTHERLAND, DOUGLAS E. DAVID B. McCLELLAN, MARK W. HULSBECK, and CHRISTOPHER M. Introduction A concern exists that wire fish traps may be too effective and may damage reef fish stocks. Current regulations in the U.S. Gulf of Mexico and South Atlantic Federal waters allow mum mesh sizes of 1 x 2″, IS’ agonal, and 1.5 xIS’. These mesh sizes retain snapper and grouper that are smaller than the minimum legal size limits and below the minimum size of first sexual maturity (Munro, 1983; Taylor and McMichael, 1983). Sutherland and Harper (1983) and Taylor and McMichael (1983) ported that 38-50 percent of the fish captured in traps had no direct mercial importance. Noncommercial species and undersized commercial fishes incur injury and mortality from: ABSTRACT-Catch and mesh selectivity of wire-meshed fish traps were tested for eleven different mesh sizes ranging from 13 X 13 mm (0.5 x 0.5″) to 76 x 152 mm (3 X 6″). A total of 1,8IOfish (757 kg) senting 85 species and 28 families were captured during 330 trap hauls off eastern Florida from December 1986 to July 1988. Mesh size significantly affected catches. The 1.5″ hexagonal mesh caught the most fish by number, weight, and value. Catches tended to decline as meshes got smaller or larger. Individual fish size increased with larger meshes. Laboratory mesh retention experiments showed tionships between mesh shape and size and individual retention for snapper janidae), grouper (Serranidae), jack angidae), porgy (Sparidae), and fish (Acanthuridae). These relationships may be used to predict the effect of mesh sizes on catch rates. Because mesh size and shape greatly influenced catchability, regulating mesh size may provide a useful basis for managing the commercial trap fishery. 36 1) Attempting to escape from traps, 2) embolisms caused by changes in bient pressure as traps are lifted to the surface, 3) stress and handling at the surface before release, and 4) tors such as moray eels which enter traps and prey on fishes before the traps are hauled (Sutherland and per, 1983; Taylor and McMichael, 1983). Lost traps (ghost traps) which continue to catch fish have also been a concern, although some evidence cates that lost traps quickly become damaged and ineffective (Sutherland et aI., 1983). Determining the effects of mesh size on fish size and composition is tant for fishery management. ing trap mesh size can reduce the chances of overfishing and can mize fishery resource production by reducing juvenile and bycatch ity. Here we examine the effects of wire fish trap mesh size on the catch composition and size distribution of reef fishes off southeastern Florida. The objectives of this research were to: 1) Document the sIze distribution of individuals and species caught by different mesh sizes, 2) determine the effects of different mesh sizes on catch of target and nontarget fishes, and 3) report the selectivity of meshes so that optimum mesh sizes can be mined for management purposes based on their capacity to reduce bycatch James A. Bohnsack, David L. Sutherland, Douglas E. Harper, David B. McClellan, and Christopher M. Holt are with the Miami tory, Southeast Fisheries Center, National ine Fisheries Service, NOAA, 75 Virginia Beach Drive, Miami, FL 33149. Mark W. beck is with the NOAA Corps, Miami tory, NMFS Southeast Fisheries Center, 75 ginia Beach Drive, Miami, FL 33149. mortality and yet retain marketable fishes. Previous studies of fish trap mesh selectivity may not be entirely able to the trap fishery in southeastern U. S. waters due to differences in cies availability, abundance, and size of fish present. This study differs most importantly in that the sampled area had received relatively little trap ing effort and that more mesh sizes were tested. Fish traps have been illegal in Florida state waters to a tance of 3 n.mi. (5.6 km) from shore since 1980, and in Federal waters at depths less than 100 feet (30 m) since 1983. Most previous studies of mesh selectivity have been conducted in heavily exploited areas outside the continental United States (Olsen et aI., 1978; Stevenson and Stuart-Sharkey, 1980; Hartsuijker and Nicholson, 1981; Hartsuijker, 1982; Munro, 1983; Luckhurst and Ward, In press). In these locations detecting differences between meshes would be more cult because larger individuals were more likely to be absent. Methods Field Methods Fish traps constructed with different sizes of wire mesh were fished in depths of 7-40 m about 5-7 km east of Key Biscayne, Fla. Field studies sisted of two phases: December 1986 to July 1987 and October 1987 to July 1988. The first phase tested eight meshes (five square and three gular) measuring 0.5 x OS’ (13x 13 mm),1.5x 1.5″(38x38mm),1x 2″(25x 51mm),2x 2″(51×51 mm),2x 3″(51 x 76mm),3x 3″ Marine Fisheries Review

PAGE – 2 ============
(76x76mm), 2×4″(51x mm), and 4 x 4″ (102 x 102 Measurements were from “knot knot.” The second phase added rectangular and one mesh: 1.5 x 3″(38x76mm),3 x (76 x 152 mm), and 1.5 x 2.3″ (38 58 mm), respectively. Mesh sizes referred to in English units for venience. The hexagonal mesh is ferred to as 1.5 ” hexagonal. Mesh characteristics and measurement versions appear in Table All traps used vinyl-coated wire were rectangular, measuring mately61x71 x91cm(2’high 2.3′ wide x 3′ long). Each trap had single funnel entrance in one end terminated in a 6 x 46 cm (2.5 x vertical opening. The top and panels of the traps were constructed the tested mesh. The side and panels of all traps were constructed 1 x 2″ (25 x 51 mm) wire mesh to present the same ette and presumably the same of visual attractiveness to fish. This x 2″ mesh was the second to so that its presence did not affect sults of other tested meshes. One was constructed entirely of I x 2″ mesh, but had all inside panels with the smallest tested mesh, 0.5 0.5″ galvanized hardware cloth. The x 2″ mesh was considered the dard mesh based on its wide and usage off southern The traps were fished unbaited in trawls (strings) of four traps. Each trawl had traps attached at 50 m vals to a 250 m groundline with a crete or steel weight anchoring each end of the groundline. A subsurface or surface buoy was often attached to one end of each groundline to aid in tion and retrieval. The traps were domly attached to the groundline to prevent sampling bias and each set was fished under similar conditions of depth and bottom type to avoid founding effects on mesh size. Soak times averaged 7 days but varied siderably (range 1-19 days) due to weather factors. Lost, stolen, or aged fish traps were replaced or paired as needed, and different traps of a given mesh size were rotated into the 51(2), Table 1.-0Imenslons of trap meshes used In field studies. Width Length Area Diagonal Width Length Area Diagonal Shape (inches) (inches) (inchesf (inches) (mm) (mm) (cm2) (mm) Square 0.5 0.5 0.25 0.71 12.7 12.7 1.6 Rectangular 1 2 2 2.24 25.4 50.8 12.9 Hexagonal 1.5 2.3 2.3 2.32 38.1 58.4 22.3 Square 1.5 1.5 2.25 2.12 38.1 38.1 14.5 Rectangular 1.5 3 4.5 3.35 38.1 76.2 29.0 Square 2 2 4 2.83 50.8 50.8 25.8 Rectangular 2 3 6 3.61 50.8 76.2 38.7 Rectangular 2 4 8 4.47 50.8 101.6 51.6 Square 3 3 9 4.24 76.2 76.2 58.1 Rectangular 3 6 18 6.71 76.2 152.4 116.1 Square 4 4 16 5.66 101.6 101.6 103.2 Table 2.-Summary of fish trap catch and effort data by mesh size and region. Mesh Trap Total Catch Total Mean wt. Mean wt. Median wt. Total size hauls catch per haul weight per haul per fish per fish species (inches) (no.) (no.) (no.) (kg) (kg) (kg) (kg) (no.) 0.5 x 0.5″ 28 322 11.50 50.46 1.80 0.16 0.08 1 x 2″ 34 210 6.18 80.65 2.37 0.38 0.21 1.5 x 1.5″ 30 259 8.63 128.13 4.27 0.50 0.22 1.5 Hex” 31 396 12.77 142.24 4.59 0.36 0.20 2 x 2″ 27 153 5.67 53.98 2.00 0.35 0.24 1.5 x 3″ 31 213 6.87 84.40 2.69 0.39 0.28 2 x 3″ 31 76 2.45 73.71 2.38 0.97 0.38 2 x 4″ 27 78 2.89 59.14 2.19 0.76 0.50 3 x 3″ 29 67 2.31 40.88 1.41 0.61 0.45 4 x 4″ 33 19 0.58 25.10 0.76 1.32 1.16 3 x 6″ 29 17 0.59 18.89 0.65 1.11 0.80 Totais 330 1.810 757.58 fishing schedule. The number of hauls for an sample size was determined to methods given by Bros and (1987). Mesh sizes added in phase were fished more often in phase II obtain comparable numbers of Each captured fish was identified, weighed, and measured to the nearest millimeter of fork length. Total length, standard length, body depth, and body width were recorded for many uals. Where possible, fish were leased after measurements were made. Economic Analysis The effects of mesh size on the value of catches were analyzed based on voluntarily reported mean sale prices for each species by 30 food dealers from 6 Florida counties for May 1988 (Economics and tics Office, NMFS Southeast Fisheries Center, Miami, Fla., personal mun.). Wholesale price per pound was converted to mean price per gram and multiplied by the weight for each cies from a standardized sample of 30 trap hauls per mesh size. Prices were adjusted according to fish size for some species as commonly done in the fishery. We assigned large individuals (> 5 pounds, 2.3 kg) the highest ues, medium sizes (2-5 pounds, 0.9-2.3 kg) the lower range of values, and small sizes « 2 pounds, 0.9 kg) a standard value of $0. 50/pound ($1. lO/kg). Mesh Retention Experiments The largest mesh that would retain a particular fish was determined during laboratory and field trials. Most of the fish used in laboratory studies were captured in fish traps during field studies, although some were obtained from other sources. Fish were tested by gently pushing them through ious meshes (beginning with the largest and proceeding to smaller meshes) until they were retained. The hexagonal mesh was not tested cause it easily became distorted during testing. 37

PAGE – 3 ============
–Table 3.-Parcent catch composition by family. Famil-Ies are listed according to decreasing percentages of 1980 weights. Data for 1980 are from the commercial trap fishery In Dade and Broward Counties (Sutherland and Harper, 1983). Data for 1987-88 are from eastern Florida (Dade County) using 1 x 2″ mesh traps only. Sample size: 3,011 kg (5,984 fish) In 1980 and 757 kg (1,810 fish) In 1987-88. Weight Numbers Family 1980 1987-88 1980 1987-88 Lutjanidae (snappers) 21.0 5.9 15.2 3.8 Serranidae (groupers) 10.0 21.0 0.9 2.4 Balistidae (Ieatherjackets) 10.0 17.0 14.8 22.4 Haemulidae (grunts) 9.1 17.0 19.2 31.4 Pomacanthidae (angelfishes) 7.5 3.7 4.7 1.9 Sparidae (porgies) 7.4 0.8 7.8 1.0 Labridae (wrasses) 6.9 1.7 3.2 1.9 Acanthuridae (surgeonfishes) 6.7 6.7 7.6 11.9 Scaridae (parrotfishes) 4.9 3.7 4.3 3.8 Ostraciidae (boxfishes) 3.3 1.1 5.1 1.4 Carangidae fjacks) 3.2 2.5 2.5 2.4 Pricanthidae (bigeyes) 2.2 0.0 3.1 0.0 Diodontidae (porcupinefishes) 1.5 1.6 3.2 3.3 Scorpaenidae (scorpionfishes) 1.4 2.0 1.8 2.4 Muraenidae (morays) 1.3 6.6 0.4 1.4 Holocentridae (squirrelfishes) 1.1 1.3 1.7 2.4 Results Catches Fish trap catch and effort data were summarized for field studies (Table 2). A total of 1,810 fish, representing 85 species in 28 families and weighing 757 kg, were captured during 330 trap hauls. The relative percent tion of various families to total catch was compared with previous data from commercial trap catches for ern Florida (Table 3). This comparison reflects only data from I X 2″ meshed traps, the predominant commercially used trap in 1980. A 1979-80 survey of commercial trap catches off Dade and Broward counties showed that snapper, grouper, triggerfish, and grunts, in decreasing order of dance, dominated commercial trap catches (Sutherland and Harper, 1983). The 1987-88 catches were dominated by grouper, triggerfish, and grunts, with snapper ranking 6th in weight. In the current study, mean catches ranged from a low of 0.58 fish/haul for a 4 X 41/ mesh to 12.77 fish/haul for the IS’ hexagonal mesh (Tables 2, 4; Fig. 1). With the exception of the 0.5 x OS’ mesh (which had the second highest average catch in numbers) the 38 20 15N u M B E R 10 F I S H 5 0 MEAN NO. FISH PER TRAP HAUL ttt i+ ..-r-+ .5x.5 lx2 1.5sq 1.5 hex 2×2 1.5×3 2×3 2×4 3×3 4×4 3×6 MESH SIZE (INCHES) MEAN WEIGHT PER TRAP HAUL 7.,————————-, 6 5 K I 4 L o G R A M S 11 . o+——————-.L.j -1 -‘—,—r——r—-,–,–,——,r——r—-,–,—.—-‘ .5x.5 1×2 1.5sq 1.5 hex 2×2 1.5×3 2×3 2×4 3×3 4×4 3×6 MESH SIZE (INCHES) MEAN WEIGHT PER FISH 1.6 1.4 t1.2Kt++tI L 0 G 0.8 R A 0.6 M S 0.4 ++ ++ 0.2 + 0 .5x.5 1×2 1.5sq 1.5 hex 2×2 1.5×3 2×3 2×4 3×3 4×4 3×6 MESH SIZE (INCHES) I I 95% CI -£-Mean I Figure I.-Effects of mesh size on fish trap catches. Bars show means and 95 percent confidence intervals. Sample sizes are in Table 2. average number of fish per haul tended to decline with meshes larger or smaller than 1.5″ hexagonal. The total number of species caught in larger mesh traps was considerably less than with smaller mesh (Table 2). Mean total weight per haul tended to decline with meshes larger or smaller than 1.5″ hexagonal, ranging from a lowof0.65 kgfor a3 x 6″ meshtoa high of 4.59 kg for the 1.5″ hexagonal mesh (Tables 2, 5; Fig. 1). Mean weight per fish tended to crease with mesh size, especially for Marine Fisheries Review

PAGE – 4 ============
Table 4.-Welght (g) offish caught by various meshes off southeastern Florida between December 1986 and July 1988. Weight of fish caught (g) by mesh size (inches)’ Species 0.5xO.5 lx2 1.5xl.5 1.5 Hex 2×2 1.5×3 2×3 2×4 3×3 4×4 3×6 Total Acanthurus behianus 595 1,905 100 2,810 724 1,705 225 8.064 Acanthurus chirurgus Acanthurus coeruleus 245 1,486 2,005 4,553 5,555 3,315 4,791 7,607 8,330 865 7,039 3,893 1,245 6.976 2,860 36,405 24,360 A/uterus schoep” Aluterus scriptus Anisotremus surinamensis 9,766 507 2,910 600 4.’;10 675 475 11,072 605 1,252 4,425 17,206 2,175 1,020 4.063 450 54,142 2,387 5,182 Anisotremus virginicu5 Aulostomus maculatus 450 200 2,019 2,289 5,915 1,505 4,645 16,823 200 Balistes capriseus Balistes vetufa 425 970 377 2,057 391 1,482 4,220 1,482 Calamus bajonado 475 495 1,220 2,190 Calamus calamus 625 309 970 1,155 3,059 Calamus proridens 765 380 400 1,545 Cantherhines macrocerus 780 780 Cantherhines pullus 99 128 375 680 1,282 Canthidermis sufflamen 145 145 Caranx bartholomaei 675 21,010 16,805 320 6,539 1,425 6,500 53,274 Caranx cIYsos 975 375 695 784 2,829 Caranx latus 121 900 1,021 Caranx ruber 1,040 380 1,420 Chaetodipterus laber 405 1,955 2,360 Chaetodon capistratus 100 125 500 145 47 917 Chaetodon ocellatus 125 235 1,755 800 669 370 3,954 Chaetodon sedentarius 175 435 610 Chaetodon striatus 50 187 237 Chi/omycterus schoep” 329 329 Dasyatis americana 2,140 2,140 Diodon holoeanthus 137 1,326 770 1,025 879 922 5,059 Epinephelus morio 1,685 4,200 3,895 919 10,699 Epinephelus sp. 1,600 1,600 Epinephe/us striatus 1,380 1,380 Equetus acuminatus 97 113 210 Ginglyostoma cirratum 3,600 2,920 9,380 13,000 28,900 Gymnothorax lunebris 6,500 3,780 2,550 9,240 11,700 33,770 Gymnothorax moringa 1,574 1,574 Haemulon a/bum 415 798 1,213 Haemufon aurolineatum 13,411 200 1,337 14,948 Haemu/on carbonarium 250 250 Haemu/on flavolineatum 1,750 1,820 3,055 6,765 201 13,591 Haemulon parrai 645 615 2,738 361 1,998 485 400 7,242 Haemu/on p/umieri 4,861 6,664 5,080 11,718 8,471 3,435 40,229 Haemulon sc;urus 387 1,220 3,117 4,724 Halichoeres bivitlatus 46 46 Ho/acanthus bermudensis 1,575 5,860 1,174 1,150 7,176 10,848 3,800 7,480 39,063 Holacanthus ci/iaris 400 248 745 540 317 2,250 Ho/acanthus tricolor 19 740 500 260 250 1,769 Holocentru5 ascensionis 210 912 580 196 1,898 Ho/oeentrus rulus 175 1,180 612 1,967 Kyphosus sectatrix 800 800 Lachnolaimus maiximus 4,435 1,340 6,730 3,892 2,979 2,970 450 3,322 2,080 1,400 29,598 LactophIYs bicaudalis 400 120 520 LactophIYs polygonia 111 1,241 510 1,862 LactophfYS quadricornis 300 1,015 3,111 1,896 1,348 580 1,206 9,456 LactophIYs trigonus 591 591 LactophIYs triqueter 900 485 372 207 1,964 Luljanus analis 1,573 3,020 12,200 6,241 5,545 1,460 11,900 2,700 44,639 Luljanus apadus 2,660 2,660 Luljanus cyanopterus 7,250 6,500 13,750 LUljanus griseus 1,255 1,520 665 3,440 Luljanus joeu 766 766 Luljanus synagris 104 2,100 1,723 105 4,032 Monacanthus hispidus 587 3,462 2,291 1,975 2,948 990 1,510 250 14,013 Mulloidichthys martinicus 417 417 Mycteroperca bonaci 7,350 5,800 8,750 7,850 29,750 Mycteroperca micro/epis 13,480 4,650 18,130 Deyurus chIYsurus 395 1,375 4,880 1,680 8,330 Pomacanthus arcuatus 980 1,200 3,755 2,765 2,285 1,257 3,565 4,479 5,680 25,966 Pomacanthus paru 1,572 1,293 1,700 810 5,375 Priacanthus arenatus 285 336 621 Prionotus roseus 38 38 Pseudupeneus maculatus 2,041 428 200 255 2,924 Rachycentron canadum 1,450 1,450 Scarus coeru/eus 975 2,625 3,600 Scarus taenipterus 585 585 Scorpaena plumieri 1,649 400 270 605 500 3,424 Seriola dumeri/i 8,500 8,500 Seriola rivoliana 450 282 540 340 1,612 Sparisom chIYsopterurn 850 2,197 2,008 300 480 1,879 7,714 Sparisoma aurolrenatum 160 200 360 Sparisoma sp. 500 500 Sparisoma viride 350 570 405 1,140 5,467 460 4,520 520 700 14,132 Sphoeroides spengleri 724 724 Sphyraena barracuda 620 31,680 13,710 3,900 49,910 Umbrina coroides 90 90 Ur%phus jamaicensis 695 695 Total weight (g) 50,463 80,649 128,126 142,244 53,978 83,505 73,708 59,141 40,875 25,098 18,887 756,674 Count 85 Number of samples 28 34 30 31 27 31 31 27 29 33 29 ‘The following five partially decomposed fish were caught in the indicated meshes but were not weighed: Gymnothorax lunebris (1.5 x 1.5″) Sparisoma viride (2 x 3″); Ho/acanthus bermudensis, Pomacanthus arcuatus, and Sphyraena barracuda (3 x 3″).

PAGE – 5 ============
meshes 2 x 3″ and larger (Fig. 1, 2; Table 2). The average weight per fish ranged from a low of 0.16 kg for a 0.5 x 0.5″ mesh to a high of 1.3 kg for the 4 x 4″ mesh. Because the weight/ frequency distributions were strongly skewed (Fig. 2), median fish size was also examined. Median size increased with mesh size, ranging from a low of 0.08kgfor a0.5 x OS’ mesh to a maximum of 1.16 kg for a 4 x 4″ mesh (Table 2, Fig. 2). Median weights of fish remained relatively constant kg) for five ler meshes ranging in size from 1 x 2″ to 1.5 x 3″. Median weights of fish caught in traps with meshes of 2 x 3″ or larger were about two to five times higher (0.38-1.3 kg) than those from the five smallest mesh sizes. The effect of mesh size on ual weight was determined using a one-way analysis of variance on transformed data. The null hypothesis of no difference between mesh sizes was rejected (F = 84.50; df = 10, 1794; P < 0.05). An a posteriori least-significant difference test (LSD test) compared all possible pairs of mean weights by mesh size. Forty-five of the 55 paired mean weights differed significantly (p < 0.05, LSD test) by mesh size (Table 6). The ten paired catches that did not differ significantly tended to be for adjacent mesh size. Economics The value of catches was examined based on market categories (Tables 7, 8). Primary commercial species had the highest market value and included snappers (Lutjanidae), groupers ranidae), and hogfish, Lachnolaimus maximus. Secondary commercial cies had about half the market value of primary commercial species and cluded grunts (Haemulidae), porgies (Sparidae), triggerfishes (Balistidae), and some jacks, Seriola sp. Other species had limited or no market value. Primary commercial species were the major component of total value for most meshes although the relative contribution varied ably (Fig. 3). The estimated commercial sale value, based on a standardized sample of 30 trap hauls per mesh, 40 1000 Mesh -0.5" x 0.5" 10 No. Fish' 322 1 No. Trap Hauls· 28 0.1 I !II" I jI'iiiii, ! I '-.11 I'," ,T " 1, , '''' b.Lu-,dtl,l, I, , , 1000 If 100 1.5 x 1.5" 10LuJ 259 1 30 0.1 , 1.5" Hex 39610 31 0.1 ,,---.-CJi!','--'---1'---.----i--,1000 i 100 i X 2"x2" 153() .llllIuu 7 C o. wm, " 1nJ1CJi!"c.lc"---,---,----,---,-,-+1-,--,----.-,__I 2 1000 :::J 15" Q) , , , , , , , ,Jllltl.'''-rlr,'---'---1+-,1L. 1000 u.. 100i MD 2"x3" 1°LJu'IIi 76 , ",' ,BUdl)I, , I I, 1000 100 10 2" x 4" 781 270.1 1000 MD 3"x3" LdlJihibu.. ---I 1000 i 100iii M.D 4" x 4" 10 =I 19 o ;: ," I,," 1000Ii 3" x 6" _ 17 o. .,,--,---r-+I'l'-'--1,-,,---t--'I L.,-,---,----r--+--r-..Ł1_-,--.-29_ o 250 500 750 1000 6000 11000 Weight class (g) Figure 2.-Weight-frequency of trapped fish by mesh size. Fish ing less than 1,000 g were grouped into 50 g intervals. Fish. weighing more than 1,000 g were grouped into 1,000 g intervals. X is mean weight, MD is median weight. ranged from $O.4l/haul for the 4 x 4" meshes smaller and larger than the mesh to $5.42/haul for the IS' 1.5" hexagonal mesh and was roughly gonal mesh (Fig. 3). Catch value, correlated to total numbers and weight though variable, tended to decrease for per haul (Fig. 4). Marine Fisheries Review PAGE - 6 ============ Table 5.-Specles and number of fish trapped by mesh size off southeastern Florlds between December 1986 snd July 1988. Table 6.-Dlfferences In mean fish weight as a function of mesh size.Number of fish trapped by mesh size (inches) 1 Species 0.5xO.5 1x2 1.5x 1.5 1.5 Hex 2x2 1.5x3 2x3 2x4 3x3 4x4 3x6 Total Mesh size (inches) 10 157 40 0.5 223 4Acanthurus behianus 3 1 3 1 Mesh 1.51.52 3 size xx x 1.5 x x xxxxx Acanthurus chirurgus 1 6 13 13 2128 10 17 109 9 96 (inches) 0.5 2 1.5Hex 2 3 3 4 3 4 6 Aluterus schoepfi 18 10 121 836 46 105 Aluterus scriptus 1 22 6, Acanthurus coeruleus 91820 431 5 0.5xO.5 10 lx2 Anisotremus virginicus 78 25 518 64 Anisotremus surinamensis 63 1.5Xl.5 21 1.5 Hex nAulostomus maculatus Balis/es capriscus 3 26 13 2x2 Balistes vetula 22 1.5x3 Calamus bajonado 2 4 2x3 Calamus calamus 2 34 10 2x4 Calamus proridens 21 4 3x3 4x4Cantherhines macrocerus 11 12 3x6 Can/hidermis sufi/amen 1 Caranx bartholomaei 22192 2 38 ,. = significant difference (p < 0.05, LSD test). Caranx crysos 512 9 Can/hemines pullus 37 2n = no significant difference (p > 0.05). Caranx latus 23 Caranx ruber 34 Chaetodip/erus faber 1 2 3 Table 7.-Wholesale market value, based on voluntary Chae/odon capistratus 3 11 2 1 18 reports by 30 dealers from six Florida counties for May Chaetodon ocellatus 315 10 64 39 1988. Chae/odon seden/arius 17 8 Chaetodon striatus 1 2 3 Number of Price ($/Ib) Chifomycterus schoepfi 1 dealers Dasyatis americana 1 1 Category reporting Mean Min. Max. Diodon ho/ocanthus 74 5 46 27 Epinephelus morio 2322 9 13 0.38 0.20 0.50 Epinephelus sp. 1 1 Angelfish 1 0.15 0.15 0.15 Epinephelus s/riatus 1 1 Sailfish 9 0.24 0.05 0.60 Equetus acuminatus 2 1 3 Grouper, black 21 2.05 1.40 2.40 Ginglyostoma cirratum 2 2 2 7 Grouper, gag 15 1.96 1.40 2.30 Gymnothorax funebris 1 1 1 6 Grouper, Nassau 5 1.65 1.45 2.00 Gymnothorax moringa 2 2 Grouper, red 19 1.63 1.15 2.20 Haemulon album 2 Grouper, scamp 13 2.21 1.70 2.80 Haemulon aurolineatum 179 2 15 196 Grouper, snowy 8 1.76 1.45 2.20 Haemu/on carbonarium 1 1 Grouper, Warsaw 6 1.30 0.90 1.90 Haemulon flavolinea/um 17 15 21 52 1 106 Grouper, yellowedge 3 1.83 1.60 2.00 Haemulon parrai 2 2 9 1 7 2 24 Grouper, yellowfin 4 1.85 1.60 2.00 HaemuJon plumier; 38 36 26 64 40 13 217 Grouper, other, mixed 5 1.78 1.65 2.20 HaemuJon sciurus 3 7 17 27 Grunts 12 0.36 0.20 0.60 Halichoeres bivittatus 1 1 Hagfish 6 1.33 1.00 1.50 Holacanthus bermudensis 3 6 2 1 8 12 3 6 41 Jacks, crevalle 13 0.29 0.20 0.70 Holacan/hus cifiaris 1 2 1 1 6 Rays 2 0.06 0.05 0.06 Holacanthus tricolor 4 2 1 1 9 Snapper, lane 8 1.04 0.65 1.50 Holocentrus ascensionis 4 2 1 8 Snapper, mangrove 17 1.44 1.00 2.25 Holocentrus rufus 1 7 4 12 Snapper, mutton 10 1.77 1.50 2.05 Kyphosus secta/rix 1 1 Snapper, yellowtail 8 1.86 1.50 2.40 LachnoJaimus maiximus 2 4 12 13 6 6 4 3 52 Snapper, other, mixed 4 1.99 1.80 2.10 Laclophrys bicaudalis 1 1 2 Triggerfish 8 0.71 0.50 1.05 Lactophrys polygonia 1 3 2 6 Porgy (White snapper) 11 0.55 0.30 0.80 Lactophrys quadricornis 4 13 9 5 2 6 40 Misc. food fish 11 0.29 0.20 0.35 Lac/ophrys /rigonus 11 Laclophrys trique/er 31 1 16 Lutjanus analis 21 6 4 9232 29 Lutjanus apodus 10 10 LutjanU5 cyanopterus 2 Mesh Retention Luljanus griseus 42 3 9 Lutjanus jocu 1 Luljanus synagris 1159 1 26 A total of 758 fish among 62 species Monacanthus hispidus 42814 12218 8 96 were tested to determine their ability to Mulloidichthys martinicus 22 Mycleroperca bonaci 11 4 escape mesh of different sizes. The Myeteroperca microlepis 21 3 Ocyurus chrysurus 25 214 32 largest mesh able to retain a fish was Pomacanthus arcuatus 12 5 4 3584 36 determined for the six most common Pomacan/hus paru 2 227 Priacanthu5 arenatus 2 families (Table 9). The size of retained Prionotus roseus 11 19 fish for each family was related toPseudupeneus maculatus 15 2 Rachycentron canadum 11 mesh size and shape. Because of bio-Scarus coeruleus 45 Scarus taenip/erus 33 logical variability, different fish of the Scorpaena plumieri 5 210 same species and length mayor maySeriola dumerili Seriola rivoliana 111 4 not be retained by a particular Sparisom chrysop/erum 465 4 21 Because area changes exponentially Sparisoma aurofrena/um 31 4 Sparisoma sp.5 5 with the liner mesh dimensions, we Sparisoma viride 1 210 26 examined the relative effects of meshSphoeroides spengleri 20 20 Sphyraena barracuda 15 2210 size on numbers, weight, and value Umbrina coroides 11 per haul based on the area of the mesh Urolophus jamaicensis 1 opening (Fig. 4). Total number 101 210 259 396 153 213 76 78 67 19 17 1,810 Number of samples 28 34 30 31 27 31 31 27 29 33 29 41’The number of species reported differs from Table 4 because some fish were counted but not weighted.

PAGE – 8 ============
——– Table S.-Contlnued. Value ($) by mesh size (inches) Species P.S,O 0.5xO.5 lx2 1.5×1.5 1.5 Hex 2×2 1.5×3 2×3 2×4 3×3 4×4 3×6 Total haul in the 1980 study. Catches the 1.5″ hexagonal mesh were more Mesh Area (cm251(2), 1989 Ho/ocentrus rulus 0 Kyphosus sectatrix 0 Lachnolaimus maiximus P Lactophrys bicaudalis 0 Lactophrys polygonia 0 Lactophrys quadricornis 0 Lactophrys trigonus 0 Lactophrys triqueler 0 Lutjanus analis P L. analis medium P L. analis, small P Lutjanus apodus P L. apodus, medium P L. apodus, small P Lutjanus cyanapterus P Lutjanus griseus P L. griseus, medium P L. griseus, small P Lutjanus joeu P Lutjanus synagris P L. synagris, small P Monacanthus hispidus 0 Mulloidichthys rnartinicu 0 Mycteroperca bonaci P Mycteroperca microlepis P Ocyurus chrysurus P O. chrysurus, medium P O. chrysurus, small P Pomacanlhus arcualus 0 Pomacanthus paru 0 Priacanthus arenatus 0 Prionotus roseus 0 Pseudupeneus maculalus 0 Rachycenlron canadum P SCarus coeruleus 0 SCarus laenipterus 0 Scorpaena plumieri 0 seriola dumerili S Seriola rivaliana S Sparisom chrysoplerum 0 Sparisoma aurolrenatum 0 Sparisoma sp. 0 Sparisoma viride 0 Sphoeroides spengleri 0 Sphyraena barracuda 0 Umbrina coroides 0 Urolophus jamaicensis 0 Tolal value 1Commercial classification: P = primary, S = secondary, 0 Figure 4.-Relative effects of mesh area on mean total weight, mean total numbers, and mean total value per haul. II) C) asc: II) 0 II) Comparing our results to a study of Q” the 1979-80 commercial trap fishery II) off Dade and Broward Counties > (Sutherland and Harper, 1983), we +: as found commercial species comprised Q)a: 66 percent by weight and 64 percent by number vs. 77 percent by weight and 62 percent by number in 1980. We averaged 2.4 kg and 6.2 fish per haul compared to 5.6 kg and 11.1 fish per 0 0 13.92 0 0 0 0 0 5.1 0.74 0 0 0 0 0 0 0 0 0 0 0 0.08 0.29 35.56 0 0 0 0 0 0 0 0.01 1.4 0 0 0.4 0 0 0 0.58 0.11 0.34 0.24 0 0.22 0.01 0 $79.40 olher. 25 20 15 10 5 0 0 0.08 0 3.46 0 0 0 0 0.51 10.39 0 0 0 0 0 0 1.33 0.49 0.34 0 0 0.08 0.4 0 0 51.35 0 0 0.38 0.29 0.46 0 0 0.24 3.38 0 0 0 0 0.33 1.24 0.11 0 0.32 0 0 0 0 $99.18 0.62 0 19.72 0 0 0.19 0 0 44.41 1.59 0 0 0 0 0 3.71 0.59 0 0 0 1.8 0.3 0 0 0 0 0.7 1.13 0.4 0 0 0 0 0 0 0 0 0 0.24 1.28 0 0 0.26 0 10.47 0 0 $135.01 20 0.31 0000000 $1.01 0.41 0000000 $0.41 11.03 9.7 8.42 1.27 10.81 6.3 3.73 0 $88.36 0.08 000000 $0.33 0 0 0 0.07 0.88 0.34 0 0 $1.29 0 1.92 1.17 0.96 0.38 0 0.8 $6.05 0000000 0.39 $0.39 0 0 0.3 0.23 0.15 0 0 0 $1.19 22.22 0 15.58 5.51 51.55 10.89 0 0 $165.65 0.6702.5 00 000 $5.50 00 0.1 00000 $0.10 00000000 $0.00 1.580000 00 0 $1.58 0000000 $2.05 0 0 30.75 0 31.66 0 0 0 $62.41 00000 000 $5.04 00 000000 $1.08 0.550000000 $0.89 003.2500000 $3.25 00000000 $0.00 1.4300 0.090000 $3.40 0.43 0.13 0.19 0 0.03 0 0 $1.81 00 000 000 $0.29 25.35 0 0 38.24 0 0 0 36.67 $135.82 19.430 0 0 00 00 $70.78 02.34000000 $2.34 0.652.790000 00 $4.14 4.77000 0000 $6.28 1.01 0.73 0.4 1.31 1.53 1.71 0 $8.58 0 0 0 0 0.47 0 0.51 0.28 $1.72 00.17 00.17 0 0 0 0 $0.34 0000 0000 $0.01 00.16 0 0 0 0 0$ 1.92 0 0 00 0 00 0$ 3.38 01.6200 000 $2.22 000 00000 $0.40 000000 00 $0.00 6.890000 000 $6.89 00.50.28 0 0 0 0 0 $1.35 0.34 1.16 0 0 0 0 0 $4.79 00000 000 $0.22 0000 0000 $0.34 3.88 0.28 2.79 0.37 0.46 0 0 $9.30 000000 00 $0.00 0 0 0 4.38 0 1.33 0 0 $16.40 00000 000 $0.01 000 00.1 00 0$0.10 $162.54 $42.84 $100.63 $67.11 $110.01 $35.56 $12.31 $40.85 $885.44 I-Weight -+-Numbers *’ Value $ I 40 60 80 100 120 43

PAGE – 9 ============
Table 9.-Fork length (em) of flahes retained by different trap meshes. Family and mesh size 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Fork Length (em) 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Total retained Snapper 1×2″ 1.5×1.5″ 2×2″ 2×3″ >3×3″ 1 1 6 4 2 2 1 11 4 24 7 2 23 6 2 4 1 3 2 3 6 7 4 1 1 4 1 1 2 1 3 1 3 1 152 75 28 27 11 11 Grouper 1 x2″ 2xZ· 2×3·’ >3×3″ 4 2 1 2 2 1 2 2 2 1 2 2 26 4 7 9 6 Grunts 0.5xO.5″ 1×2″ 1.5×1.5″ 2xZ’ 2×3″ 3 5 3 6 3 2 6 13 6 3 18 12 2 1 16 9 3 12 4 11 2 6 2 24 5 2 16 5 2 1 7 4 9 6 3 2 7 3 2 2 1 250 23 72 36 78 41 Jacks 1 xZ’ 1.5×1.5″ 2xZ’ 2×3″ 3 6 3 3 2 2 5 5 5 4 2 2 3 2 2 2 3 3 59 24 25 5 5 Porgy 1 xZ’ 1.5×1.5″ 1.5×3″ 2×3″ >3×3″ I2 1 1 3 1 2 2 2 1 5 2 1 3 34 8 4 4 10 8 Surgeonfish 1×2″ 1.5×1.5″ 2×2″ 2×3″ >3×3″ 3 3 2 5 4 7 8 8 2 2 8 5 3 8 2 1 3 1 5 3 1 4 7 1 2 101 24 53 23 1 0 similar to those reported during the 1980 study, averaging 4.6 kg and 12.8 individuals per haul. Differences in the present and earlier study partially flect differences in trap designs, area fished, and method of fishing. In this study we tended to sample in lower water with smaller traps which may account for the differences in catch data. Other studies have shown that catches are significantly affected by the type of funnel opening, trap size, trap shape, bait and other ables (Luckhurst and Ward, In press). Selectivity A concern of commercial fishermen is that fish will not enter traps with larger sized meshes because these traps are less visually distinctive. We found that commercial species will ter traps with a wide variety of mesh sizes. The walls of all traps were structed with 1 x 2″ wire mesh so that they presented the same visual ette and did not bias catches due to differential attraction. Luckhurst and Ward (In press) noted mesh selectivity could be biased by fish attraction to different trap silhouettes. The darker trap silhouette created by the 0.5 X 0.5″ mesh lining a 1 x 2″ mesh was apparently not more attractive to larger fish than were the other unlined traps which had a standard 1 x 2″ wall mesh. Although the 0.5 x 0.5″ trap had one of the highest catch rates by numbers (11.5 fish/haul), the mean weight/haul (1. 8 kg) was similar to those reported for much larger meshes (Fig. 1, Table 2). The high numbers in the 0.5 x 0.5″ mesh are mainly counted for by many small fishes, such as the tomtate, Haemulon neatum, that could escape all larger mesh sizes (Table 4). Other related behavioral responses that affect recruitment to traps (Hartsuiker and Nicholson, 1981) should have equally affected catches by different mesh sizes in our study. Captured fish size was mately related to trap mesh size (Fig. 2), confirming earlier studies by Olsen et al. (1978), Stevenson and Sharkey (1980), and Munro (1983). However, capture of a particular fish was not strictly a linear response to mesh size as measured by either the area of the mesh opening or the longest open dimension. Retention responses for a particular species were fluenced by mesh shape as well as the size of the opening (Table 9). Sutherland et al. (1987) showed that both fish size and body shape were important factors explaining ences in retention by a given mesh size between species. Slender (terete) fishes (e.g., eels, lizardfishes, cobia) of a given length (or weight) were much more likely to escape a partic-Marine Fisheries Review 44

PAGE – 10 ============
ular mesh than were compressed fishes (e.garable with that of the smaller sized trap function as 2 x 2″ meshed trap. angelfishes, triggerfishes, fishes) or depressed fishes (e. g., rays, flatfishes) of the same length. Rounded (fusiform) fishes fell tween the two extremes. For example, a surgeonfish with a relatively slender body shape might escape a rectangular mesh but not a square mesh of the same area. A grouper which has a more rounded body shape might be more likely to escape a square mesh than a rectangular mesh. Thus, tions of mesh size and shape aimed at optimizing one species may greatly fect capture of other species due to differences in body shape. Total value, total species caught, number of individuals, and mean total weight per haul tended to decline with meshes larger and smaller than the 1.5″ hexagonal mesh (Fig. 4). Two of the minimum mesh sizes currently legally specified (1.5 x 1.5″, and 1.5″ hexagonal) had the greatest percentage contribution to total weight and total value. Mesh sizes 2 x 3″ and larger, especially, tended to catch larger fish but fewer species and individuals. Based on these results, the presently specified legal minimum mesh sizes appear to do little to reduce bycatch. Catchability Results show that catchability (the proportion of a population removed by one unit of fishing effort) can be greatly influenced by mesh size and shape. Fewer primary commercial cies were caught with the largest mesh sizes. This reduced catch partially flects the lower availability of large fish that can be retained in large meshes. Also, fish may be less willing to enter large meshed traps, perhaps because fewer retained fish make the trap less attractive. Economics Assuming constant effort, a larger mesh size would have immediate fects on total revenue of the trap ery by lowering catchability. Larger mesh sizes would provide less revenue per trap haul. With larger mesh sizes, more effort (number of hauls) must be expended to obtain total revenue 51(2), 1989 mesh. To achieve the same revenue with larger meshes as obtained with a 1.5″ hexagonal mesh, fishermen would have to increase their number of trap hauls anywhere from 1.5 to 13 times depending on the mesh size (Fig. 3,4). The number of trap hauls men can make is limited by their skill, manpower, time, and equipment. The simple economic analysis done here is limited. It does not consider potential future benefits of allowing fish to escape and grow before ing the trap fishery, direct impacts on market prices due to supply, or ble losses to the future fishery from natural mortality. Also, price per pound is highly variable between markets and over time. These siderations are beyond the scope of this study. Mesh Retention Laboratory studies show that mesh retention depends on the species and size of the fish tested (Table 9) as well as on the mesh shape and size land, et aI., 1987; In press). These results do not consider availability in the fished area or willingness to enter traps. Laboratory tests of mesh tion on individual fish show only the physical limitation of fishes to escape a given mesh size. Quite possibly some fish passing though a given mesh in the laboratory would not, or could not, escape under actual field tions. With these qualifications, Table 9 provides a basis to estimate mesh sizes necessary to allow the escape of fishes of specific sizes for the majority of commercial species. For example, mesh size of 2 x 3″ or larger should allow snapper and grouper less than 30.5 cm (12″) to escape. Federal regulations in the Gulf of Mexico currently require four (2 x 2″) escape windows in each trap. The fects of escape windows were not cifically investigated in this study due to logistical, fiscal, and time tions. However, a conservative proximation of their effect can be tained by extrapolation of the data from 2 x 2″ meshes. In the extreme, the escape windows would make the Based on our observations of fish behavior it is likely that most fishes able to escape a 2 x 2″ opening will freely swim in and out of the escape window while the trap is resting on the bottom. However, when a trap is pulled, most fishes react by swimming toward the bottom and are unlikely to find the escape window. Thus, injury and mortality from lifting and handling are still likely to occur. These fish would be more likely to escape during lifting if the entire top and bottom panels were made of the desired escape-sized mesh similar to the trap used in the field study. An advantage of fish traps over tom longline, trawl, or hook and line fishing is the increased selectivity of fish traps based on mesh size. It is possible to fish traps with meshes that reduce the capture of fish below a imum size. Hooks are less selective for fish size; small fish can be captured on large hooks. Thus, the mortality and injury associated with lifting smaller fish off the bottom can be reduced or avoided with fish traps more easily than with hooks. Presumably sized hooked fish still face trauma from handling and embolism even if released. Summary This study has described the effects of mesh size on selectivity, retention, catchability, and value of fish trap catches. Mesh size and shape fluence trap retention. In this study the most effective mesh sizes for total enue per haul and total weight were the 1.5 x 1.5″, and 1.5″ hexagonal meshes, two legally specified imum mesh sizes. Commercial species will enter a wide variety of mesh sizes. Increasing mesh size reduces catchability and enue per haul which, within limits, can be compensated for by increasing fort (number of hauls). Adjusting mesh size offers a means for regulating and managing the reef fish fishery. Fish traps with appropriate mesh sizes potentially can reduce bycatch and undersized fish injury and mortality more effectively than similar 45

277 KB – 11 Pages