9. Discards and bycatch in Shrimp trawl fisheries.

The highest rates of incidental catch of non-target species as identified by Alverson et al (1994) and other workers is associated with shrimp trawling. The use of selection technology in reducing the incidental catch of non-shrimp in temperature waters does, however, seem to be affecting the amount of bycatch discarded and recent evidence, suggests that in many parts of the world tropical shrimp bycatch which was once discarded is now being utilised. The figures which suggest that large quantities of fish are wasted in certain regions may be over estimates because it had been assumed that non-shrimp catches from tropical shrimp fisheries was automatically equated to discards. (See Table 11)

Alverson et al for instance suggest, that in the Northwest Pacific 97% of the shrimp bycatch is discarded producing over 4 million tonnes of waste fish. This however is countered by information provided by Zhou and Yimin (1996) who suggest that the Chinese shrimp trawl fleet discards very little of the non-shrimp catch. According to Zhou and Yimin the shrimp fishery of China catches about 1.8 million tonnes of bycatch all of which is used, much for feeds for the Chinese aquaculture industry. In S E Asia there has also been a growth in recent years in industry’s which use bycatch from shrimp fisheries for human consumption (Chee 1996).

Recent (1997) evidence from the countries of central America and the Caribbean suggest that the amount of incidental catch that is now utilised in the region is greater than that suggested by Alverson et al. In countries such as Belize, Colombia, Costa Rica and Nicaragua for instance it seems that between 20 and 30% of the once discarded bycatch is now utilised and in Venezuela it is probably at least 30%. In Guyana and Brazil the amount of utilisation has increased and probably now accounts for about 10% of the incidental bycatch. In Cuba where the state operates fishing, marketing, processing and distribution it is thought that virtually all edible bycatch is now used which probably accounts for up to 70% of the non-shrimp catch. (F Teutscher, Pers comm.) These figures seem to contradict the blanket figure of 3% utilisation that Alverson et al suggested for the whole of the West Central Atlantic, and perhaps indicate fuller use of the catch in a similar fashion to the trends seen in South East Asia. See Table 11 below.

The trend for greater and more complete utilisation of bycatch has been noted over the years and as Andrew and Pepperell (1992) state "It seems likely.......that the utilisation of by-catch will increase as demand for protein escalates and the retention and processing of by-catch becomes more economically viable".

Table 11 : Estimated bycatch and discard levels from shrimp fisheries of the world

Area	Estimated Bycatch	Estimated Discard	Percent Discarded
N W Atlantic	81,665	80,031	98
N E Atlantic	210,297	206,091	98
W C Atlantic	1,310,653	1,271,334	97
E C Atlantic	123,636	61,818	50
Mediterranean/Black Sea	257,859	250,124	97
S W Atlantic	253,446	245,842	97
S E Atlantic	39,143	19,571	50
W Indian Ocean	1,871,075	748,430	40
E Indian Ocean	482,879	289,727	60
N W Pacific	4,284,408	4,155,903	97
N E Pacific	28,269	27,421	97
W C Pacific	1,450,352	1,377,835	95
E C Pacific	590,955	561,416	95
S W Pacific	19,446	18,863	97
S E Pacific	203,677	197,567	97
Total			85

Based on: Alverson et al Table 4

9.1 Temperate waters

9.1.1 Pandalus

In temperate waters there has been extensive introduction of selector devices in trawls in recent years which have enabled reduction in the catch of non-target species to be made. This work centres around the use of the Nordmøre grate which originated in Norway as a means of reducing the capture of jelly fish but it was also found that incidental catch of fish was reduced without substantial reduction in the catch of shrimp. (Isaksen 1996) The technology has been further developed in particularly Norway, Canada and Australia and incidental catches have now been reduced in the shrimp fisheries of various regions. The grate works by separating active swimming fish, which will swim out of the net to avoid passing through the grate, from passive shrimp which will pass through the grate into the cod end. The device not only greatly reduces the catch of unwanted fish but will also lead to better quality catches, because less large objects get mixed with the shrimp in the cod end and less time is required on hauling and to sort out the catch before preservation. (Løbach and Viem 1996). In addition it has been suggested that fish escaping from nets fitted with the grates have a high survival rate, particularly compared with fish which escape through the meshes of the cod end of a trawl, which can be so damaged and traumatised by the experience that there chances of long term survival are much reduced. (Vold Soldal & Engås 1997)

In the northern Pandalus borealis shrimp fishery of Canada bycatch, particularly of redfish, has been almost eliminated by the use of Nordmøre grates, and new fisheries for shrimp have been opened up in Canadian waters which were not practised before because of high bycatch rates. Bycatch has been reduced from over 20% to less than 5% of total catch in some instances. In 1993 Norway and Russia agreed to introduce compulsory use of the grates in the shrimp fishery in both their EEZ’s as well as around Svalbard and their use is mandatory from January 1997 on bottom trawls in the Barents Sea. The use of the grate is now also compulsory in many of the northern shrimp fisheries of the United States of America, Canada and Iceland.

9.1.2 Nephrops

The Nephrops norvegicus fisheries in the North Atlantic and North Sea also produce bycatch some of which is discarded. The Belgian fishery, according to Redant and Polet (1994), discards very small numbers of undersized cod, plaice and sole but there are sometimes discards of large numbers of whiting and dab, depending on area fished and season. However the fishery catches as a matter of course white fish such as cod, plaice and sole for sale when at or above legal size and gunards and dabs are also sold on local markets as well as Nephrops which is the nominal target catch. Figures for the landings (1993) of the most important fish in the catches show that Nephrops in fact only constitutes just over 35% of the landed weight and just under 50% of the value. (See Table 12 and Table 13) It appears, therefore, that the fishery does not target Nephrops exclusively and will react to market demands, auction prices and the catch rates of Nephrops, targeting other species or fishing other grounds for other species. The regulation concerning landing sizes for the various round and flat fish caught along side the Nephrops dictate or influence the amount of these species discarded. The authors estimate that discarding of undersize and unmarketable whiting by the Nephrops fishery, although amounting to large numbers per year, is small in comparison to discards of whiting from other fishing activities and that probably the discarding of undersize cod, plaice and sole are negligible in comparison with others sources of discarding. This case illustrates the problems associated with the definition of target and non-target species in fishery. The fishery in this case is called the Nephrops fishery but the figures suggest that if only Nephrops was caught and marketed it would not be viable.

Table 12 - Top ten species in the landings of the Belgian Nephrops fishery 1993 (tonnes)

Table 13 - Top ten species in the landings of the Belgian Nephrops fishery 1993 in terms of auction sales figures

BEF = Belgian Francs

This situation in the Belgian Nephrops fishery seems to contrast with others. Evans et al (1994) report that only 12% of the catch in the Nephrops fishery of the Farne Deep (North Sea) is Nephrops and that a large number of other species, both vertebrates and invertebrates are also caught. Although some lemon sole, plaice, whiting, cod, haddock and sting ray are landed the vast majority of the bycatch was juvenile fish of species of no commercial value and discarded. They conclude that most of the fish was dead before being discarded and much was scavenged by sea birds.

Thorsteinsson (1992) asserts that capture of small haddock has often been a problem in the Icelandic Nephrops fishery. A double trawl one with square mesh window of 135 mm mesh proved that the window allowed the effective release of small haddock and whiting and some cod. All Icelandic Nephrops trawlers were obliged to use such windows very soon after the work concluded.

It seems therefore that progress is being made in eliminating or very much reducing discards in shrimp fisheries in temperate waters. This is happening through the use of selection devices such as the Nordmøre grate in the North Atlantic and Australian waters or by the fuller use of fish that is caught alongside the target species. The use of square mesh panels in Nephrops trawls seems to hold some promise for reducing non-target catch and is being introduced in some fisheries. In others however the catch of non-primary target species actually make valuable contributions to the commercial viability of the fishing units.

9.2 Tropical and subtropical waters

The Nordmøre grate in its original configuration works by excluding large specimens and retaining the smaller ones which are allowed to pass into the cod end of the trawl net. In tropical shrimp fisheries the shrimp are generally part of a species assemblage where the size of the target shrimp overlap with the size of the fish that need to be excluded and it is often, in fact the smaller specimens that need to be excluded from the net. The size characteristics, therefore, of the species flock and behavioural differences between shrimp and fish mean that the direct transfer of this technology has not been possible into tropical fisheries although some work is continuing in this field with the use of a grids and other devises mounted in the top panel of tropical shrimp trawls allowing the smaller fish and shrimp to escape, showing some promise.

The public conception of the problem of bycatch in tropical shrimp fisheries is, however, different from that associated with the problem in temperate fisheries. It is a fact that many of the tropical shrimp trawling activities take place in waters adjacent to low income, food deficit countries. The pressure in this scenario is not, necessarily, to stop the bycatch being caught but to make use of it for feeding people once it has been caught. Work, for instance, on the use of grid technology in the shrimp fishery of Mozambique was focused on the reduction of catches of "only the very smallest fish..........as the intention of a newly approved project was to take care of and process most of the bycatch taken by the shrimpers" (Isaksen and Larsen 1993) In this case, in the final analysis it seems, that the Mozambique government were reluctant to introduce a top mounted grid in trawls as it reduced incidental fish catch too much. They preferred instead to look further at a larger size cod-end mesh which reduced only the catch of the very smallest fish and shrimp and retained the large valuable shrimp for export and the fish which could be used for local consumption (R Larsen, Pers comm.)

9.2.1 Species composition

An analysis of the literature reporting the species composition of bycatch in tropical shrimp trawl fisheries shows that there are a large number of species represented in the bycatch but that they represent a relatively small number of families and that some families are found in the bycatch in many parts of the world.

Table 14 below illustrates the commonality of families that occur in the bycatch of shrimp trawlers throughout the topical world. For instance Ariidae (Marine Catfishes), Carangidae (Jacks), Clupeidae (Herrings, shads, sardines, menhaden), Gerreidae (Mojarras), Sciaenidae (Croakers), Trichiuridae (Ribbon Fish) are found in bycatch in all the regions shown in the table. Other species groups such as the Leiognathidae (Pony Fish) are not so wide spread but where they are found can make up large proportions of the non-target catch.

Table 14 - Families of fish reported as tropical shrimp trawl discards/bycatch

Sources: Alverson et al (1994); Papers from: Gulland and Rothschild (1982); U Khum Mg Aye (1995); Chee (1996); Bejei (1980); Barratt (1986); Hendrickson and Griffin (1993); Gordon (1991); Feidi 1989

Andrew and Pepperell (1992) conclude that the four main characteristics of shrimp bycatch are:

1. Fin fishes make up the majority of the catch in many shrimp fisheries.
2. The sizes of the fish are generally small <20 cm and often of similar size to the shrimp.
3. A relatively small number of species of fishes may dominate the by-catch.
4. Several families of fish are abundant in the bycatch of many of the worlds shrimp fisheries (such as sciaenids, pomadasyids, sparids, synodontids, serranids, bothids, and nemipterids)

9.3 The early work on tropical shrimp bycatch use

In the 1970’s the growth of shrimp fisheries in tropical and warm waters and the perceived growth of demand for animal feed stuffs in those countries with growing shrimp fisheries led to the search for means of utilising bycatch for animal feeds. Experiments on the production of fish silages or fish hydrolysates based on methods already used in Denmark and Poland were made. (Disney et al 1978) The goal of this work was to produce a simple technology that could be used on a small scale in tropical fisheries where the main source of raw material was to be found. The use of organic and inorganic acids or the use of acid producing bacteria were investigated which, when added to a mixture of small fish, would produce an stable liquid which could be incorporated into animal feeds. This work in UK, Indonesia, Hong Kong, Sri Lanka, Australia and elsewhere suggested that it was technically feasible to produce such products and that they had potential as feeds for livestock. However the uptake of this technology in tropical shrimp fisheries has been small. In spite of the fact that the technology is simple it still requires the supply of acids or acid producing bacteria, a change in the way in which animal feeds are produced (since a liquid rather than a solid are made) and extensive training of potential users. Twenty or more years since the first work was done and there are few if any examples of its uptake in the target communities.

In the late 1970’s and early 1980’s the focus had changed in the search for answers to the utilisation of shrimp bycatch. The proceedings from a conference in Guyana in 1981 (IDRC 1982) illustrate that workers were now looking at the use of bycatch to produce new products mainly for human consumption. The products under investigation sought to disguise the original fish by the production of minced products from the flesh either as salted products (Young 1982, Bligh and Duclos 1982) or as frozen minces (Tan at al 1982, Productos Pesqueros Mexicanos 1982, Tableros and Young 1982). Other investigations used existing technology to endeavour to make products resembling existing products such as canned pates, sausages, frozen fish sticks etc. (Poulter 1982). Recent evidence suggests that very few of these technology lead initiatives have provided long term solutions to the use of bycatch with few examples of products still being made commercially.

9.4 Gulf of Mexico and South Atlantic prawn fisheries of the United States

A study conducted by the Gulf and South Atlantic Fisheries Development Foundation into bycatch reduction devises applicable for use in the Gulf of Mexico and the South Atlantic Shrimp Fisheries analysed in some depth the species composition and quantity of catch in the two fisheries. The study was prompted because of concern that shrimp fishing may have been contributing to over fishing of commercially important species through capture of juveniles of those species. Capture of juvenile red snapper (Lutjanus campechanus) in the Gulf of Mexico and in the South Atlantic of weakfish (Cynoscion regalis) along with king and Spanish mackerel (Scomberomorus cavalla and S maculatus) in both fisheries had been cause for concern and it was thought that the incidental catch of these species in the shrimp trawl fisheries was limiting the resource available for fisheries directed at these species and the main aims of the programme were to reduce the capture of these species although, as can be seen below, other less commercially important species were more common in the incidental catch.

In the Gulf of Mexico Fish Shrimp trawl fishery was concluded that over 450 taxa were identified in the catch and average catch was about 27 kg/hour of trawling. The catch could be divided into three major components as shrimp 26%, other invertebrates 17% and finfish 67% giving a ratio of 5.25 bycatch to 1 of shrimp. This figure contrasts the figure of 10.3:1 given in Alverson et al for the Gulf of Mexico shrimp fishery.

The top ten species found in catch were as follows in Table 15

Table 15 - Top ten species found in Gulf of Mexico shrimp trawl catches

Species - English Name	Percentage of Total Catch
Longspined porgy	15
Brown shrimp	9
Atlantic croaker	9
Inshore lizardfish	6
Pink shrimp	3
Gulf butterfish	3
Lesser blue crab	2
White shrimp	2
Longspined swimming crab	2
Brown rock shrimp	2

In the South Atlantic about 150 taxa were found in shrimp trawls with a catch rate of about 26 kg/hour. Shrimp made up 18%, other invertebrates 31% and finfish about 51% of the total catch. This gives a bycatch to other species ratio of about 4.5 to 1. The top ten species in the catch are listed as follows in Table 16.

Table 16- Top ten species found in South Atlantic shrimp trawl catches

Species - English name	Percentage of Total Catch
Cannonball jellyfish	14
White shrimp	9
Spot	9
Atlantic menhaden	9
Brown shrimp	8
Other jellyfish	8
Atlantic croaker	6
Southern kingfish	4
Blue crab	4
Star drum	3

9.5 Northern Prawn Fishery of Australia

The area of sea within Australia’s EEZ from Cape York in the east to Cape Talbot in the west is a rich area for capture of penaeid prawns. Known in Australia as the Northern Prawn Fishery (NPF) it yields quantities of the shoaling banana prawns (Penaeus merguensis) which involve little if any bycatch and quantities of the more disperse tiger prawns (P. esculentus and P. semisulcatus) which involve capture of considerable quantities of non-target fish or bycatch. A study by the Northern Territory Department of Primary Industry and Fisheries of bycatch between 1987 and 1992 (Pender et al 1992a) looked at the likely utilisation options for the common species in the bycatch. They concluded that 97% of the bycatch was discarded at sea with presently only valuable species being retained, these are detailed in Table 17.

Table 17 - Components of the bycatch in the northern prawn fishery retained onboard

Of the discarded bycatch most was fish, with over 200 taxa recorded, belonging to 75 families. The major families of the discarded fish bycatch are given in Table 18 which shows that they are primarily of families which have a small size at maturity and were currently un-used in Australia.

Table 18 - Major families of fish discarded in the mixed species fishery in 1988

The studies identified a number of species with potential commercial value within the discarded catch. (Pender et al 1992b). The options for use included fish for direct human consumption, production of surimi, use as bait, fertilisers and animal feeds. The major city and capital of Northern Territories - Darwin, has groups of residents from Asian and other back grounds for whom some of the fish in the discarded bycatch are more familiar, than for Australians. A number of species in the bycatch were assessed for their acceptability by these groups and assessments of the viability of utilisation made. The data indicated that taking into account the population sizes and the amount of fish potentially available the following had market potential - trevallies, goatfish, threadfin breams and emperors. Available in smaller quantities but with higher demand from certain groups are species such as black pomfret, Bombay duck, whiting, tufted sole and octopus. The findings also indicated that generally any fish more than 13-15 cm in length could be acceptable for consumption.

The work also looked at the possibility of production of surimi, canned, dried and frozen products. It was concluded that, although feasible technically, there would be, at best, marginal profits on such enterprises.

Non-food use as bait, animal feeds and fertilisers were also investigated. Squids, pilchards, garfish, mullets and cuttlefish have potential as bait for anglers and commercial fishermen. The aquaculture of barramundi (Lates spp) requires supplies of raw fish. Three species in the bycatch were found particularly suitable as food for barramundi, with others which were tried being less acceptable and other species not eaten at all by barramundi. See Table 19.

Table 19 - Fish species preferences of Lates calcarifer brood stock