Steven J. Kennelly

Optimal positioning and design of behavioural-type by-catch reduction devices involving square-mesh panels in penaeid prawn-trawl codends

Two experiments were done in the oceanic penaeid prawn-trawl fishery in New South Wales to investigate (i) the optimal positioning of behavioural-type by-catch reduction devices (BRDs) involving square-mesh panels; and (ii) the relative performances of two industry-developed square-mesh panels against a composite square-mesh panel developed by scientists. In the first experiment, three codends, each containing one square-mesh panel (located on the tops of the codend at distances of 0.7, 1.2 and 1.6 m anterior to the last row of meshes respectively) significantly reduced the catches of some small fish and total discards compared with a control codend that held no BRD. Rates of reduction significantly increased with proximity of the square-mesh panel to the end of the codend. There was, however, a significant reduction in catches of prawns from the codend containing the square-mesh panel at 0.7 m. Based on these results, the optimal position for these sorts of BRDs was determined to be at 1.2 m anterior to the last row of meshes. In the second experiment, at this position, two codends containing industry-developed plastic and metal square-mesh panels were less effective in excluding by-catch than a codend containing the composite square-mesh panel. The results are discussed in terms of species-specific differences in behavioural responses and swimming ability and the future development and testing of behavioural-type BRDs in penaeid prawn trawls.

Strategies for Improving the Selectivity of Fishing Gears

Few fishing methods and gears are entirely selective for the targeted species and their sizes. The majority of gears have incidental catches (collectively termed ‘bycatch’) that vary from isolated occurrences in some hook-and-line fisheries to large numbers of juveniles of key species in trawl fisheries. Of primary concern is the contribution that the mortalities of such bycatches may have on subsequent stocks. Over the past 20 years, extensive efforts have been directed towards addressing this issue by modifying problematic fishing gears (especially trawls) and practices. Whilst this work has facilitated considerable reductions in bycatches (up to 80% in some cases), very few (if any) of the changes made to existing gears are 100% effective. There remains, therefore, a substantial mortality of unwanted individuals in most fisheries. To work more comprehensively towards the ultimate goal of achieving perfect selectivity, we propose that, in addition to conventional methods used in recent decades to modify fishing gears, a more lateral approach should also be adopted involving completely alternative gears. Specifically, we propose a strategy that: (1) examines the boundaries of what is realistically achievable in modifying poorly selective gears using established bycatch reduction protocols; and (2) determines the utility of alternative gears that, because of their design and/or operation, have selective mechanisms which could be applied to problematic gears. In this paper, the logic involved in the first approach is discussed and data supporting the benefits of the second approach are presented. Reducing the discarding of small prawns Project No. 2001/031 88 NSW Dept of Primary Industries The issue of bycatch From the earliest evidence of fishing more than 90 000 years ago (Yellen et al. 1995) to the present day, humans have exponentially advanced their harvesting methods. The clear focus of these developments has been to maximize the catches of an ever-increasing diversity of targeted species, with little or no regard for the incidental catches (termed ‘bycatch’, sensu Saila 1983). A progression from simple harpoons, hooks and traps deployed from the shore, through nets set from boats, to the industrial factory trawlers of developed countries has culminated in technology which, in many cases, far exceeds the sustainability of local resources. This excess was evident at the end of the 20 century by the collapse of many commercially-important stocks, a plateau in the world’s total landed wild catch (at less than 100 million tonnes) and the volumes of bycatch discarded in pursuit of targeted catches (Alverson et al. 1994). While recognition of the potentially negative impacts of unchecked fishing technology date back to the 14 century (Dyson 1977), it is only during the last few decades that coordinated attempts have been directed towards improving the selectivity of fishing gears (Kennelly and Broadhurst 2002). Relevant reviews of the published literature suggest that nearly all fishing gears and methods have received at least some attention (e.g. gillnets – Hamley 1975; longlines – Løkkeborg and Bjordal 1992; traps – Mahon and Hunte 2001), although the majority of effort has been directed towards benthic trawl fisheries (e.g. Kennelly 1995) and especially those targeting shrimp (Andrew and Pepperell 1992; Broadhurst 2000). This has occurred in response to the disproportional ratio of retained-to-discarded catches and the amount of unwanted catch discarded each year by shrimp trawlers; estimated to represent between 30 and 60% of the total world harvest of wild fisheries resources (Alverson et al. 1994). While the absolute volume of bycatch associated with shrimp trawling clearly makes it one of the most the most problematic fishing methods, many other gears including fish trawls, seines, gillnets, traps and longlines have, in recent times, been identified as having significant selectivity issues and have consequently been associated with prolonged calls for improvements coming from a variety of environmental groups, recreational fishers, interacting commercial fisheries and the general public. Solving bycatch problems During the past 2 decades, problems surrounding the issue of bycatch has shifted the focus of fishing gear technology from catching as much of the target species as possible (with little regard for collateral impacts) to improving selectivity, both in terms of the species targeted and their desired sizes (Kennelly and Broadhurst 2002). In many cases, the successful development and adoption of solutions to improve selection in problematic gears can be summarized in a simple framework (see also Kennelly and Broadhurst 1996; Kennelly 1997; Broadhurst 2000) which involves industry and researchers each applying their respective areas of expertise to the particular problem. This framework comprises five key steps: (1) quantifying bycatches (mostly via observer programs), (2) identifying the main bycatch species and their sizes of concern, (3) developing alterations to existing fishing gears and practices that minimize the mortality of these species, (4) testing these alternatives in appropriately-designed field experiments and (5) gaining acceptance of the new technology throughout the particular fishery and interested stakeholders. The protocol for completing the framework is quite straightforward and has been described with numerous examples by Kennelly and Broadhurst (1996), Kennelly (1997) and Broadhurst (2000). The crucial and most difficult step (3 above) is the actual development of appropriate solutions that improve the selectivity of existing fishing gears for the targeted catch and so reduce unwanted bycatch. Depending on the type of gear and its particular problems, solutions may involve simple adjustments to operational procedures and/or existing components of the gear, like changing the size and/or shape of meshes or hooks. Alternatively, for many towed gears, more complicated modifications that include physical bycatch reduction devices (BRDs) may need to be invented or modified from other fisheries (Broadhurst 2000). Owing to their relative complexity, these types of modifications frequently require detailed adjustment and reassessment to exclude specific sizes of individuals or species, yet maintain targeted catches (Kennelly and Broadhurst 2002). Project No. 2001/031 Reducing the discarding of small prawns NSW Dept of Primary Industries 89 While the above framework summarizes several successful attempts at addressing the problems of bycatch in different fisheries throughout the world (Kennelly 1997), in many cases the established protocols for improving inherently problematic gears has restricted fishing technologists in terms of working towards the ultimate goal of perfect selectivity. A reason for this is that to ensure the industry adoption and acceptance of modified designs that reduce bycatch (i.e. step 5 above), nearly all researchers have aimed to achieve 100% retention of the targeted species (during step 3). Theoretically, it should be possible to dramatically improve the selectivity of most fishing gears, provided some concomitant sacrifice in their overall efficiency is permitted. The issue would then become what is an acceptable loss of the targeted catch in order to improve selectivity and reduce bycatch. An extreme solution for achieving ‘perfect selectivity’ may be to reorder the above logic and, using traditional gears and established bycatch reduction methods, approach a 100% exclusion rate of unwanted catch at any cost to the desired catch. This approach could be appropriate in tightly-regulated fisheries where there is imminent threat of closure due to discarding. Reductions in gear efficiency could also be offset via some compensatory increases in the value of the targeted catch through ‘eco-labeling’. This sort of strategy would not be feasible, however, in the vast majority of countries and especially those where artisanal fisheries represent the main source of income for communities. For these fisheries, bycatch reduction clearly needs to be maximized with minimal impact on the efficiency of the gear for the targeted catch. Maximizing gear development within existing bycatch reduction frameworks To approach maximum bycatch reduction with no loss of the targeted catch (during step 3 of the framework described above), there needs to be a general estimate of what is achievable for particular gears. As a starting point, this requires an assessment of the limits of established modifications for improving selectivity. For many conventional towed gears, different sizes and/or shapes of mesh are among the simplest alterations and their utility is often (or at least should be) defined first. Under the framework proposed by Broadhurst (2000), this involves testing beyond what might intuitively be appropriate, so that the limits of a particular range of mesh sizes or shapes can be quantified and defined. If the solution to reducing particular bycatch species of concern is not apparent within the boundaries of the simple alterations tested, then more complex modifications (including physical BRDs) warrant examination. Specific designs of BRDs should also be tested to define their limits. For example, if mechanical-sorting grids are required to exclude organisms larger than the targeted species, then a range of configurations that include very narrow and wide bar spacings and small and larger profiles or angles of orientation should be examined (e.g. Broadhurst et al. 2004b). Similarly, because factors like relative water flow strongly influence the performance of BRDs that operate by exploiting differences in the behavior of species (Broadhurst et al. 1999a), these sorts of modifications need to be tested at different positions throughout the gear (e.g. Broadhurst et al. 2002). Coherent hypotheses encompassing the full range of key factors influencing the performance of mod

Using composite square-mesh panels and the Nordmøre-grid to reduce bycatch in the Shark Bay prawn-trawl fishery, Western Australia

Abstract An industry-modified Nordmore-grid was tested on its own and with a composite square-mesh panel as a secondary bycatch reduction device (BRD) at two different positions (aft and forward) in codends. Compared to the control codend (which had no BRDs), all three combinations (Nordmore-grid only, Nordmore-grid and aft composite square-mesh panel and Nordmore-grid and forward composite square-mesh panel) reduced catches of prawns (mostly western king prawns, Penaeus latisulcatus). However, prawns were found to escape out of the Nordmore-grid—not through the composite square-mesh panels. The Nordmore-grid with the aft composite square-mesh panel significantly reduced the weight of bycatch (by 49%) and the numbers and weights of several commercially and non-commercially important bycatch species (by up to 75.7%). No other significant differences were detected. The results are discussed in terms of the likely factors influencing the performance of the various designs, including the behaviour of fish in codends, influences of hydrodynamics on their escape and the importance of the positioning of BRDs in codends.

Simulated escape of juvenile sand whiting (Sillago ciliata, Cuvier) through square-meshes: Effects on scale-loss and survival

Abstract Two laboratory experiments were done to assess effects of simulated escape through square-meshes on the scale-loss and survival of (i) non-fatigued and (ii) fatigued small sand whiting (Sillago ciliata). In experiment 1, non-fatigued fish that were forced through square-meshes (treatment fish) showed no significant difference in scale-loss compared to fish that did not pass through square-meshes (control fish), although there was a 50% difference in mean scale-loss immediately posterior to their maximum height. In experiment 2, fish were fatigued to exhaustion by swimming against a current of 0.7 to 0.8 knots for 15 min. Fatigued fish that were then forced through square-meshes showed significantly more scale-loss across their entire body than did the fatigued control fish (difference in means of between 67% to 84%). In both experiments the total scale-loss on treatment fish was quite low (1.4–4%) and there were negligible mortalities (only 2 treatment fish died in experiment 1) over the duration of each experiment (30 days). We conclude that the composite square-mesh panel currently used to reduce by-catch in the NSW oceanic prawn trawl fishery is likely to cause negligible damage and mortality of small sand whiting.

Fate of juvenile school prawns, Metapenaeus macleayi, after simulated capture and escape from trawls

Two laboratory experiments were done to assess the fate of juvenile school prawns, Metapenaeus macleayi, after simulated multiple capture and escape from trawls. In the first experiment, prawns that were trawled and escaped one, five or 10 times, sustained some physical damage (mostly limited to the loss of antennae), but this was not significantly different from that sustained by control prawns that had not been trawled. Similarly, there were no significant differences between the different treatments and control prawns in their stress levels (as measured by changes in concentrations of L-lactate). Levels of L-lactate were greatest in all prawns immediately after the experiment started and then significantly reduced after 24 and 48 h. In the second experiment, treated prawns were trawled and escaped 10 times and then monitored for mortalities over 2 weeks. Compared with control prawns (that were not trawled), significantly more treated prawns died at the end of the 2 weeks, but the overall post-trawl survival rate was >89%. It is concluded that the multiple contact and escape of juvenile school prawns from trawls had minimal effect on their overall condition.

Bycatches of endangered, threatened and protected species in marine fisheries

Bycatch remains one of the most significant fisheries issues in the world and its monitoring and reporting is now expected in many regions. This paper provides a global synthesis of the data that are available on one of the most controversial components of bycatch, that associated with the capture and discarding of endangered, threatened and protected (ETP) species in marine commercial and artisanal fisheries. We examine the available literature regarding estimates for the key taxa in this category of bycatch (seabirds, turtles, sea snakes, marine mammals, sharks, rays and teleosts) and use the data to try to provide a total global estimate. We estimate (albeit quite imprecisely) that at least 20 million individuals of such species are discarded annually throughout the world. However, there remain far too many gaps and uncertainties across fisheries and regions in the information to provide any robustness (or variance) around such an estimate, nor to determine the actual fates of these animals (many may survive). This is exacerbated because: (1) the occurrences of such species are often rare and controversial and so go either unnoticed and/or unrecorded; (2) different levels of protection are afforded to different ETP species in different countries and fisheries and; (3) discarding practices vary greatly across a hierarchy of spatio-temporal scales and according to individual fishing conditions and procedures—the latter affecting actual mortalities. Nevertheless, there have been major initiatives established in recent years to provide better data on such interactions in addition to novel fishing methods and practices that reduce them and also improve the survival of discarded individuals. This paper discusses the data currently available and the quite significant gaps that remain.