Vous êtes sur la page 1sur 8

Fisheries Research 45 (2000) 18

Assessment and benets of damage reduction in prawns due to use of different bycatch reduction devices in the Gulf of Carpentaria, Australia
John Salinia,*, David Brewera, Margaret Farmera, Nick Rawlinsonb
b

CSIRO Marine Laboratories, PO Box 120, Cleveland, Qld 4163, Australia Australian Maritime College, PO Box 21, Beaconseld, Tasmania 7250, Australia

Received 16 April 1999; received in revised form 2 September 1999; accepted 12 September 1999

Abstract The use of bycatch reduction devices (BRDs) and turtle exclusion devices (TEDs) will be compulsory in the Northern Prawn Fishery of Australia by the year 2000. These devices may also benet shers by excluding large animals that can damage prawns in the trawl and thereby improve catch quality, and hence value. This study measures the reduction in physical damage to prawns caught with BRD-tted nets compared to a standard codend. The measure of damage was that a prawn had one or more of the types of damage that, in commercial shers, would condemn it to a less valuable ``soft and broken'' category. The change in prawn damage ratio was between 6.1% and 34.7% depending on the BRD being used. The total catch weight and weight of large animals (>5 kg) were signicantly reduced with a grid (excluder device) in the net. The best BRD for retention of prawns and damage reduction was the Super Shooter Fisheye. The economic value of reducing prawn damage was calculated from a combination of the percentage reduction in damage and the price difference between 10 kg ``soft and broken'' prawns and higher-priced, nger-laid prawns in 1.5, 3.0 or 5.0 kg packs. The increased value of catches from reduced prawn damage in two BRDs that did not signicantly affect prawn catches was conservatively estimated to be $735 per week, which is more than the cost of the BRD. # 2000 Elsevier Science B.V. All rights reserved.
Keywords: Bycatch reduction devices; Prawns; Damage

1. Introduction The effect of prawn-trawl sheries on bycatch populations is of world-wide concern (Saila, 1983; Pender and Willing, 1989; Andrew and Pepperell,

Corresponding author. Tel.: 61-738267244; fax: 61738267222. E-mail address: john.salini@marine.csiro.au (J. Salini).

1992; Kennelly, 1995). Recent studies in Australia have looked at ways to reduce the bycatch from prawn-trawl catches (Broadhurst and Kennelly, 1994, 1995; Broadhurst et al., 1996; Brewer et al., 1998). In northern Australia, research has focussed on excluding large and endangered species such as turtles (Mounsey et al., 1995; Robins-Troeger et al., 1995) and on sh (Brewer et al., 1998). In Australia's Northern Prawn Fishery, night-time trawling directed at tiger prawns (mainly Penaeus esculentus and P. semi-

0165-7836/00/$ see front matter # 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 7 8 3 6 ( 9 9 ) 0 0 1 0 4 - 6

J. Salini et al. / Fisheries Research 45 (2000) 18

sulcatus) kills large amounts of bycatch. To reduce this bycatch, turtle excluder devices (TEDs) and bycatch reduction devices (BRDs) will be compulsory on prawn-trawl nets in the shery from 2000. The ratio of bycatch to target prawn species ranges from 8:1 to 21:1 in the Northern Prawn Fishery (Pender et al., 1992); the ratio during the trawling reported here from the Albatross Bay region of the shery was about 16:1 (Brewer et al., 1998). Brewer et al. (1996, 1998) tested several different exclusion grids in combination with escape windows such as large square-mesh windows, sheyes and radial escape sections. Although the use of BRDs will soon be compulsory for all shers licensed to trawl for prawns in the Northern Prawn Fishery, their successful introduction will be made easier if there is no signicant prawn loss or decline in catch value. Prawn quality determines the price of tiger prawns in the export market, for which most prawns are packed in nger-laid packs of 1.5, 3.0 and 5.0 kg packs instead of loose-packed size-graded 10 kg packs. Damaged prawns make up the bulk of such 10 kg packs (personal communications from Raptis & Newshing Aust). One seafood company trawling for deepwater prawns off eastern Australia has used turtle excluder devices (TEDs) to increase their product value by eliminating sharks, rays and other deepwater large animals from catches. Prior to the use of TEDs, bycatch induced damage had greatly reduced the value of entire shots of these delicate prawns (Sandy Wood-Meredith, Wood Fisheries, Queensland, Australia, personal communications). Any BRD introduced into the shery should ideally improve catch quality as well as successfully excluding turtles, sharks, rays, and other bycatch species. Reducing bycatch especially of larger animals, may also reduce damage to the tiger prawns, which would result in an increase in catch value attributable to the BRD. Previous studies on damage to bycatch in the Australian prawn sheries have focussed on the damage to commercially valuable nsh (Broadhurst et al., 1996, 1997) and nsh bycatch generally (Farmer et al., 1998). One aim of this study was to compare the numbers of damaged prawns that had been caught in standard codends and in codends with BRDs installed. The other aim was to calculate the economic benet resulting from a decrease in damage to prawns sufcient to offset any loss in prawn catch. Discarded

nsh are not discussed as they have no commercial value in the Northern Prawn Fishery. 2. Methods and materials 2.1. Study site The trawl-gear experiments were conducted in commercial prawn-trawl grounds west of Duyfken Point near Weipa in the north eastern Gulf of Carpentaria during October/November 1995. This was in the middle of the shing season when up to 14 commercial trawlers were in the vicinity during the 26 nights of experimental trawling. Water depth was in the range of 1525 m. 2.2. Sampling gear and design Our sample design was of balanced semi-systematic incomplete blocks where each trawl comprised a ``block'' with two ``units'' (Brewer et al., 1998). All trawls were of 2 h duration with two devices towed simultaneously (port and starboard sides). During the rst half of the sampling, six devices (AusTED, Nordmore grid Fisheye, Fisheye, Nordmore grid Square-mesh Window, Super Shooter Fisheye, and Control) were tested. To account for any port/starboard net differences, each device was used on either side an equal number of times. This required 15 tows to complete all possible BRD combinations. During the second half of the sampling, the AusTED and the Fisheye were eliminated from the comparisons due to their unacceptable loss of commercial prawns, and the four remaining devices Nordmore grid Fisheye2, Nordmore grid Square-mesh Window2, Super Shooter Fisheye2 and the Control net) were rotated through all possible pairings every six tows. There were, therefore, different numbers of stations or tows for each gear type, with the Control net towed the most number of stations, as it stayed the same throughout the sampling. The BRDs plus the Control net were used in a total of 80 paired trawls to make simultaneous comparisons of the devices against the Control codend and against each other. Eight BRDs were tested: Australian turtle exclusion device (AusTED), Fisheye, Nordmore grid Fisheye, Nordmore grid Fisheye2, Nordmore grid Square-

J. Salini et al. / Fisheries Research 45 (2000) 18

mesh Window, Nordmore grid Square-mesh Window2, Super Shooter Fisheye, Super Shooter Fisheye2 and a Control (currently used codend with no BRD). All BRDs except the Fisheye consisted of a metal excluder grid (Nordmore, Supershooter and AusTED) with an opening through which large animals are deected, combined with a sh-escape device (Fisheye and Square-mesh Window). The technical details of each BRD are described in detail in Brewer et al. (1998). Three BRDs were modied slightly during the sampling: Nordmore grid Fisheye became Nordmore grid Fisheye2, Nordmore grid Square-mesh Window became Nordmore grid Square-mesh Window2 and Super Shooter Fisheye became Super Shooter Fisheye2 (Brewer et al., 1998). The catch in each net was sorted on board to species, weighed, counted and the data entered onto an on-board computer with an Oracle database. 2.3. Damage measurements All commercial species of penaeid prawns were sampled (maximum of 50 prawns per species per trawl) for damage to produce a damage score (Table 1) based on the presence (1) or absence (0) of each of the eight damage parameters summed over each prawn. Damage score varied from 0 to 8, undamaged prawns had a score of 0. The damage parameters listed in Table 1 are the criteria used by shers to exclude

prawns from nger-laid 1.5, 3.0 or 5.0 kg packs and relegate them to the ``soft and broken'' bulk 10 kg packs. Additional data were recorded on moult stage (soft/hard), antennal damage/loss, missing eyes, mouthparts broken, swimming and walking legs missing or broken, but were not included in the analysed data as they do not condemn the prawn to the ``soft and broken'' category. Newly moulted prawns were excluded. From the damage score, a value for damage per prawn was calculated as the total damage score divided by the number of prawns sampled (Tables 1 and 2). The value is an indication of the relative level of prawn damage incurred with each BRD. The number of damaged prawns divided by the sample size produced a damage ratio per BRD, and was used to assess the reduction in damage to tiger prawns due to BRDs. A percentage change in damage ratio was calculated for each BRD relative to the Control codend (Tables 2 and 3). This differs from the change in damage listed in Table 4, which is the arithmetic difference in damage ratio, expressed as a percentage, for each BRD compared to the Control codend. The smallest loss of commercially important prawns recorded was a 10% loss by the Nordmore grid Fisheye which represents up to 20 kg of prawns valued at up to $500. As this completely swamps any economic gain in value of prawns due to reduced damage, this value is not presented in Table 4. The economic signicance of a reduction in damage is assessed

Table 1 Prawn body parts examined for trawl-induced damage and the two measures of damage used, damage per prawn and damage ratio. Damage was recorded as a presence or absence (0 or 1) ranging from 0 to 8 per prawn. Soft prawns (recently moulted) and antennal loss were recorded but not used in the analyses Body part Head Description Head missing Head loose Rostrum broken Exoskeleton broken Carapace totally crushed Telson missing Exoskeleton broken Uropods missing/broken Score 0/1 0/1 0/1 0/1 0/1 0/1 0/1 0/1 Damage score Sum of scores/sample size Total damaged/sample size

Carapace Tail

Total (sum) Mean damage per prawn Damage ratio per BRD

J. Salini et al. / Fisheries Research 45 (2000) 18

Table 2 Damage information for all commercial species for each BRD tested are presented below. Damage per prawn was recorded for the same criteria in Table 1. Damage ratio: number of damaged prawns divided by number of prawns sampled per gear; damage score per prawn: sum of the total damage score (Table 1) divided by the number of prawns sampled per gear. Change in damage ratio: the percentage change compared to the Control Gear Stations (n) Prawns caught (n) 1649 3780 1656 1685 2070 1643 2132 1675 2170 18460 Prawns damaged (n) 201 557 223 161 254 214 294 209 277 2390 Damage ratio (SE) 0.122 0.147 0.135 0.096 0.123 0.130 0.138 0.125 0.128 (0.09) (0.06) (0.09) (0.08) (0.07) (0.09) (0.08) (0.10) (0.07) Change in damage ratio (%) 17.0 8.2 34.7 16.3 11.6 6.1 15.0 12.9 Damage Score 268 796 296 240 336 289 402 277 362 3266 Per prawn 0.163 0.211 0.179 0.142 0.162 0.176 0.189 0.165 0.167 0.177

AusTED Control Fisheye NordFEye NordFEye2 NordSMWin NordSMWin2 SshootFEye SshootFEye2 Total

13 32 13 13 21 14 21 12 20 159

0.129 (0.03)

Table 3 Catch statistics and mean damage ratios for tiger prawns (P. esculentus and P. semisulcatus) for each BRD tested are presented below. The damage criteria are listed in Table 1. Catch wt.: total weight of the catch including large animals (>5 kg), but not turtles; damage ratio: number of damaged (any damage criterion) prawns divided by sample size. Change in damage ratio: the percentage change compared to the Control Gear Control NordSMWin2 AusTED SshootFEye2 Fisheye NordSMWin NordFEye2 SshootFEye NordFEye Total Total catch wt. 387.4 293.3 246.7 308.7 362.0 185.6 285.5 279.6 278.4 Mean wt. large animals (kg) 18.96 0 1.68 0 12.40 0 0 0.80 0 Prawns sampled (n) 2886 1566 1186 1605 1115 1320 1481 1379 1108 13646 Prawns damaged (n) 458 213 160 215 147 168 184 169 104 1818 Mean damage ratio (SE) 0.159 0.136 0.135 0.134 0.132 0.127 0.124 0.123 0.094 (0.02) (0.02) (0.01) (0.02) (0.01) (0.02) (0.02) (0.02) (0.01) Change in damage ratio (%) 14.5 15.1 15.7 17.0 20.1 22.0 22.6 40.9

0.133 (0.02)

Table 4 Changes in catch value attributable to a reduction in prawn damage. The is the change in tiger prawn catch in the trawls from this study (Fig. 6 in Brewer et al., 1998); ``decrease in damage'' is derived as the difference between damage ratios of the BRDs and Control codends expressed as %; $ increase/night is an average tiger prawn catch worth approximately $4000 per night of finger-laid tiger prawns in 1.5 and 3.0 kg packs Catch (%) AusTED Fisheyea NordFEye NordFEye2 NordSMWin NordSMWin2 SshootFEyea SshootFEye2
a b

Decrease in damage 0.024 0.027 0.065 0.035 0.032 0.023 0.036 0.025

$ Increase/nightb Loss $79 Loss Loss Loss Loss $105 Loss

23.1 1.9 10.0 23.1 57.5 23.7 10.0 10.6

Non-significant tiger prawn loss. Ignores change in tiger prawn catch.

J. Salini et al. / Fisheries Research 45 (2000) 18

for tiger prawns (P. esculentus and P. semisulcatus) only. The damage ratios were estimated for all prawn species (Table 2) and for tiger prawns alone (P. esculentus P. semisulcatus, Table 3). The mean number of prawns sampled and the mean number of prawns damaged are calculated per BRD (Table 3). The total catch weight includes large animals (>5 kg), but does not include turtle weights. 2.4. Data analysis The numbers of damaged prawns per trawl was assumed to follow a binomial distribution, with the total number of prawns sampled as the population parameter. A logistic link function was applied to the proportion parameter of the binomial distribution. Effects of factors such as ``Bycatch Reduction Device'', ``bycatch weight'' and ``weight of large animals'' were incorporated in the analysis. The possible overdispersion in the data was also taken into account by scaling the Pearson 2-value. The analysis was completed with the generalised linear PROC GENMOD (SAS 6.12), which accounted for codend pairings used in the sample design. Models 14 used the GENMOD procedure and binomial link function with the number damaged and the number sampled as dependant variables. Model 1 tested the impact of each BRD on damage after adjusting for ``side'' (independent variable). Models 24 used all BRDs with grids as one group (grids) and Fisheye as a separate BRD (independent variables). Model 2 tested the impact of grids and Fisheye on damage after accounting for ``side''. Model 3 tested the impact of grids and Fisheye on damage after accounting for ``side'' and catch weight. Model 4 tested the impact of grids and Fisheye on damage after accounting for ``side'', catch weight and weight of large animals. Models 57 used the generalised linear model (GLM) procedure. Model 5 tested the impact of all BRDs on damage ratio after adjusting for ``side''. Model 6 tested the impact of all BRDs on damage ratio after adjusting for ``side'' and catch weight. Model 7 tested the impact of all BRDs on damage ratio after adjusting for ``side'', catch weight and weight of large animals. The independent variable BRD is a group of BRDs that include a grid in

their construction. The Fisheye does not contain a grid. 2.5. Assessing economic benefit Information on prices paid to vessels for various grades of 10 kg bulk packs containing damaged prawns, and for 1.5 and 3 kg nger-laid packs of prawns were obtained from two of Australia's largest seafood processors and exporters (A. Raptis & Sons Pty Ltd., Colmslie, Queensland, and Newshing Australia Pty Ltd., Osborne Park, Western Australia). Tiger prawns packed in value-added 1.5 or 3 kg packs sold for between $15 and $28 per kg depending on size, and damaged prawns sold for between $8 and $11 per kg (1996 prices, A. Raptis & Sons, Colmslie and Newshing Australia, Perth). A conservative average price difference of $15 per kg between damaged and undamaged prawns was used in the calculation of economic gain due to reduced damage (Table 6). This price difference was used with the calculated reduction in damaged prawns attributable to the BRDs to produce a dollar value for the reduction in damaged prawns. The gain in dollar value was modied by any loss in prawn catch caused by the BRD. 3. Results A total of 80 paired trawls representing 159 codend trawls were obtained during the 26 nights of continuous trawling. One codend trawl catch was discarded due to a torn net. A total of 18 460 prawns was checked for damage, most (13 646) of which were tiger prawns (P. esculentus and P. semisulcatus). A total of 2390 prawns was damaged to some extent (Table 2); this included 1818 tiger prawns (Table 3). 3.1. Extent of damage Most of the damage sustained by all species of prawns was in the head region (Fig. 1). The Damage Measure per prawn (Table 2) clearly shows the control codend has the highest levels of damage, closely followed by the Fisheye. The Fisheye BRD was also closest to the Control in catch weight and weight of large animals. The AusTED and Super Shooter

J. Salini et al. / Fisheries Research 45 (2000) 18

Fig. 1. Mean damage per prawn (sum of the damage score for the head and tail, divided by the sample size) in tiger prawns caught by nets fitted with BRD and the control codend. AusTED: Australian turtle excluder device; NordFE: Nordmore grid with Fisheye; NordFE2: Nordmore grid with Fisheye2 (different orientation); NordSMW: Nordmore grid with Square-mesh Window panel in topside posterior to the grid; NordSMW2: Square-mesh Window panel placed further back than in NordSMW; SShootFE: Super Shooter with Fisheye; SShootFE2: Supershooter with Fisheye (different orientation). For detailed description of these devices see Brewer et al. (1998).

Fisheye2 had the lowest levels of damage. Mean damage ratios per BRD follow a similar trend to damage-score-per-prawn, with the highest ratio in the Control codend (0.211) and the lowest in the Nordmore grid Fisheye (0.096) (Table 2). A similar pattern occurred for tiger prawns, (P. esculentus and P. semisulcatus) alone. Mean damage ratios per BRD for tiger prawns were highest for the Control codend (0.159) and lowest for the Nordmore grid Fisheye BRD (0.094) (Table 3). Few large animals were retained by the BRDs except in the case of the Fisheye (12.4 kg per trawl) BRD. The Fisheye does not have an excluder grid. Total catch weight was highest for the Control codend (387.4 kg) and lowest for the Nordmore

Square-mesh Window (185.6 kg) (Table 3) that is, all the BRDs reduced catch weight. The Nordmore Square-mesh Window not only excludes large animals, but allows swimming bycatch to escape through the window. For all commercial species and tiger prawns, the percentage change in Damage Ratio compared to the Control codend was lowest with the Nordmore Square-mesh Window2 and highest with the Nordmore Fisheye. Overall, the percentage changes in Damage Ratio were higher for the tiger prawns (Table 3) than for all commercial prawns (Table 2). The decrease in damaged prawns, measured as the difference between BRDs and the Control codend, was between 0.023 and 0.065 (Table 4). Only two of the eight BRDs the Fisheye and the Super Shooter Fisheye had a non-signicant impact on prawn catch (Brewer et al., 1998) and reduce the tiger prawn damage ratio by 17.0% and 22.6% (Table 3). The reduced damage due to these two BRDs, although not statistically signicant (Table 5), is interpreted as a potential increase in economic return to the trawler. The rst model in the analyses showed that all the BRDs reduced tiger prawn damage ratios compared to the Control (Table 3) but only the Nordmore grid Fisheye BRD reduction was signicant. However, there was a signicant decrease in damage ratio when any grid device was present in the codend (Table 5, Model 5). The signicant difference in damage ratios due to grids and Fisheye BRD groups (Model 5) disappeared when catch weight (Table 5, Model 6) and weight of large animals (Table 5, Model 7) was added to the model. Total catch weight and weight of large animals increased the damage measured by both criteria but the increase was not statistically signicant (Table 5, Models 6 and 7).

Table 5 Intercept values for the SAS GENMOD procedure with the number of tiger prawns damaged and the numbers sampled as dependant variables and BRD, side (Model 5), catch weight (Model 6) and weight of large animals (Model 7) as independent variables Model 5 6 7 P < 0.05. P < 0.01. *** P < 0.001.
** *

Intercept 0.14 2.07 2.01

Fisheye 0.02* 0.16 0.13

Grids 0.02* 0.11 0.06

Side 0.01* 0.13 0.14

Catch wt. 0.0007*** 0.0006***

Wt. of large animals

0.003**

J. Salini et al. / Fisheries Research 45 (2000) 18 Table 6 Calculation of increased value in an average 200 kg tiger prawn catch from the best-performed BRD, the Super Shooter Fisheye. A conservative average price difference of $15 per kg between damaged and undamaged prawns was used in the calculation of economic gain due to reduced damage Value added $25/kg kg Control codend Best BRD Change Gain 170 177 $ 4250 4425 175 105 Soft and broken $10/kg kg 30 23 $ 300 230 70

The increase in the tiger prawn catch ( catch in Table 4) with the Fisheye and for the Super Shooter Fisheye BRDs was 1.9% and 10%, respectively, but for the sake of the calculation, the tiger prawn catch is assumed to be the same as for a Control codend (i.e. no loss of tiger prawns). A nightly catch of $4250 of tiger prawns was used in the calculations, but the relative values can easily be transformed for night catches of any value. If only $2125 of tiger prawns is caught per night at particular times of the season, then the benet per night would be half that for $4250 per night. The combination of no loss of tiger prawns and reduced damage increased the value by $105, or 2.3%, per night for the Super Shooter Fisheye (Table 6). 4. Discussion During the planning stages of this study, we assumed that more tiger prawns are damaged when there is a large catch in the codend than when there is a smaller catch, and that the presence of large animals increased damage. Our results show that catch weight alone signicantly increases the proportion of damaged tiger prawns (Table 5). The weight of large animals, even though representing a small fraction of the total catch, also appears to increase tiger prawn damage, although the effect was not statistically signicant. Unlike smaller bycatch species, large animals can remain active in trawl catches for long periods. They may physically damage the prawn catch during the tow or during hauling up of the net. Previous video studies on the behaviour of bycatch in the codends of prawn and sh trawl nets showed clear evidence of large animals thrashing in the codend (CSIRO, unpub-

lished data). If the proportion of large animals in the catch increased, the increase in tiger prawn damage would become statistically signicant. Contrasted to this, most of the small nsh bycatch quickly tired and fell back passively into the codend catch. BRDs can effectively reduce the amount of sh bycatch and large animals caught in prawn trawls (Broadhurst and Kennelly, 1994; Brewer et al., 1996, 1998; Broadhurst et al., 1997) and hence can also affect the reduction in tiger prawn damage (and so, improve catch value). However, various BRDs perform differently in different ecosystems (Broadhurst and Kennelly, 1997; Brewer et al., 1998). The nancial benets of the best-performing BRDs would have to be calculated empirically. However, the overall principle that BRDs reduce unwanted bycatch, in particular larger animals, and so indirectly reduce tiger prawn damage is expected to be valid for most of the Northern Prawn Fishery grounds. The proportion of tiger prawns that is damaged in our experimental trawls was about double (Control Damage Ratio 0.159, Table 3) the estimates taken (measured as ``soft and broken'') from shers' logbooks (damaged ratio of about 0.07, CSIRO, unpublished data). Two things may explain this difference. Firstly, O'Neill et al. (1999) found that shers tend to over-value prawns when sorting at sea, so logbook data tends to overestimate prawn quality. A consequence to shers is that prawn processors sometimes down-grade their catch quality as recorded in logbooks, which signicantly reduces their value. Secondly, our study area normally has the highest shbycatch ratios in the Northern Prawn Fishery (Brewer et al., 1998), possibly contributing to the high incidence of damaged prawns in our experimental trawls. Consequently, the higher levels of prawn damage measured in this study, may well reect the levels of prawn damage in commercial catches. This study is the rst to demonstrate an economic benet to prawn shers from using BRD. We show that all BRDs, including the Fisheye (which does not have an excluder grid), reduced tiger prawn damage to some extent. However, this was not the sole determinant of nancial gain to the shers as some BRDs caused signicantly lower catches of tiger prawns. Our results show there is an increase in nancial return to shers using BRDs that decrease prawn damage but do not reduce prawn catch (e.g. Super Shooter

J. Salini et al. / Fisheries Research 45 (2000) 18

Fisheye). This nancial benet results from improved quality of catch by having a lower proportion of damaged prawns. Other studies have also shown that BRDs can increase the value of the catch by increasing prawn catch rates (Broadhurst and Kennelly, 1997; Broadhurst et al., 1997; Brewer et al., 1998). The signicance of reducing tiger prawn damage is borne out by the price difference between value-added 1.5 or 3 kg packs of undamaged prawns and 10 kg bulk packed damaged prawns (A. Raptis & Sons Pty Ltd., Brisbane and Newshing Australia Pty Ltd., Perth). Our conservative estimate of increased return per night by using the Super Shooter Fisheye is $105 (2.3%), or $735 per week (Table 6) compared to the cost of installing BRDs (about $500 for a pair of nets; Julie Robins, QDPI Fisheries, Deception Bay). Over the entire NPF eet, this small increase in prawn value extrapolates to about $1 million per year. 5. Conclusion From the year 2000, all trawlers in the Northern Prawn Fishery will be required to use devices that reduce the amount of unwanted bycatch. Fishers' concerns for using BRDs are mainly based on perceived loss of nancial returns. All BRDs tested would improve the rate of nancial return if based on the reduction in damage alone, but not all maintained prawn catches. Our results show that the Super Shooter Fisheye and the Fisheye BRDs successfully achieved both goals: reduced damage without loss of tiger prawns. We have demonstrated that by using such devices, shers can increase prots of their operation. This is a strong incentive for shers to use these devices in any prawn trawl sheries that take large amounts of bycatch. Acknowledgements David Die and Janet Bishop contributed useful comments on the paper. Julie Robins, Anders Cormie, Steve Eayrs, Samantha Miller and Yougan Wang assisted on the cruise. We appreciate the cooperation and help of the crew of the RV Southern Surveyor. Yougan Wang processed the SAS analyses. The project was funded by FRDC grant No. 93/179.

References
Andrew, N.L., Pepperell, J.G., 1992. The bycatch of shrimp trawl fisheries. Oceanogr. Marine Biol.: Ann. Rev. 30, 527565. Brewer, D.T., Eayrs, S.J., Rawlinson, N.J.F., Salini, J.P., Farmer, M., Blaber, S.J.M., Ramm, D.C., Cartwright, I., Poiner, I.R., 1996. Recent advancements in environmentally friendly trawl gear research in Australia. Second World Fisheries Congress, Brisbane, Australia. Brewer, D.T., Rawlinson, N., Eayrs, S., Burridge, C., 1998. An assessment of bycatch reduction devices in a tropical Australian prawn trawl fishery. Fish. Res. 36, 195215. Broadhurst, M.K., Kennelly, S.J., 1994. Reducing the bycatch of juvenile fish (mulloway) in the Hawkesbury River prawn-trawl fishery using square-mesh panels in codends. Fish. Res. 19, 321331. Broadhurst, M.K., Kennelly, S.J., 1995. A trouser-trawl experiment to assess codends that exclude juvenile mulloway (Argyrosomus hololepidotus) in the Hawkesbury River prawn-trawl fishery. Marine Freshwater Res. 46(6), 953958. Broadhurst, M.K., Kennelly, S.J., 1997. The composite squaremesh panel: a modification to codends for reducing unwanted bycatch and increasing catches of prawns throughout the New South Wales oceanic prawn-trawl fishery. Fish. Bull. 95, 653 664. Broadhurst, M.K., Kennelly, S.J., O'Doherty, G., 1996. Effects of square-mesh panels in codends and of haulback delay on bycatch reduction in the oceanic prawn-trawl fishery of New South Wales, Australia. Fish. Bull. 94, 412422. Broadhurst, M.K., Kennelly, S.J., Watson, J.W., Workman, I.K., 1997. Evaluations of the Nordmore grid and secondary bycatch-reducing devices (BRDs) in the Hunter River prawntrawl fishery, Australia. Fish. Bull. 95, 210219. Farmer, M.J., Brewer, D.T., Blaber, S.J.M., 1998. Damage to selected species escaping from prawn trawl codends: a comparison between square-mesh and diamond-mesh. Fish. Res. 38, 7381. Kennelly, S.J., 1995. The issue of bycatch in Australia's demersal trawl fisheries. Rev. Fish Biol. Fish. 23, 165174. Mounsey, R.P., Baulch, G.A., Buckworth, R.C., 1995. Development of a trawl efficiency device (TED) for Australian prawn fisheries. I. The AusTED design. Fish. Res. 22, 99105. O'Neill, M.F., Die, D.J., Taylor, B.T., Faddy, M.J., 1999. Accuracy of at-sea commercial size grading of tiger prawns in the Australian Northern Prawn Fishery. Fish. Bull. 97, 396401. Pender, P.J., Willing, R.S., 1989. Trash or treasure? Aust. Fish. 48, 3536. Pender, P.J., Willing, R.S., Cann, B., 1992. NPR bycatch a valuable resource? Aust. Fish. 51(2), 3031. Robins-Troeger, J.B., Buckworth, R.C., Dredge, M.C.L., 1995. Development of a trawl efficiency device (TED) for Australian prawn fisheries. II. Field evaluations of the AusTED. Fish. Res. 22, 107117. Saila, S.B., 1983. Importance and assessment of discards in commercial fisheries. FAO Fishery Circular No. 765, 62 pp.

Vous aimerez peut-être aussi