bycatch and discards
In this chapter, an overview of bycatch and discard issues for some U.S. fisheries in the Northwest Atlantic is provided, with emphasis on results from quantitative stock assessments that have included discard estimates. Management efforts aimed at reducing bycatch and discards are considered. Finally, a brief review of information on the impacts of discards on some common stock assessment calculations is provided.
Fisheries in the Northwest Atlantic are generally well developed, and most resources are either fully exploited or overexploited. The diversity of directed fisheries results in a high degree of interaction among fisheries. Bycatch, as elsewhere, presents both problems and opportunities for the fisheries of the Northwest Atlantic. The fact the region's resources are so highly mixed and diverse has resulted in very few fisheries truly focused on single-species targets. As an example, even for the dredge fishery for scallops, finfish bycatch of species, such as summer (Paralichthys dentatus) and yellowtail flounders (Limanda ferruginea) and monkfish, makes up important economic elements of the unsorted catch. With the recent decline in scallop populations, revenue from the finfish bycatch from this fishery--once sold for cash by the crew as “shack landings”--has come to represent an important overall source of vessel income. Likewise, in years when highly valued species, such as haddock, cod, and yellowtail flounder, were abundant, trawl tow times were short and directed fisheries generated catches yielding high proportions of these target species.
The significant decline in traditional groundfish species over the last decade (NOAA/NMFS 1993) has resulted in much less species-specificity in targeting. One consequence of the decline in groundfish is longer tow times in trawl fisheries, now averaging about three hours. These long tow times mean heterogeneous bottom habitats are sampled during the tow, so the diversity of species caught increases as the dominance of one or more “targets” declines, and in all likelihood survival of the animals in the catch also declines. Species once considered non-target or “bycatch” have become components targeted by generic trawl effort. Therefore, the concept of “retained bycatch” has given way to target “species mixtures” in recent years in this fishery.
Discard problems of the region's fisheries are manifested in several important ways: (1) marketable species too small or otherwise prohibited from landings are nevertheless retained in the gear and are subsequently discarded, (2) species for which no current market exists are caught simultaneously with commercial or recreational species and are discarded, (3) various fleet sectors compete directly for landings or indirectly through bycatch interactions, and (4) non-fishery bycatch species, including marine mammals, turtles, and birds, are inadvertently killed. These issues are elaborated in the discussion which follows.
The discard of sublegal sized fish is significant in many fisheries, and stems primarily from the incompatibility of minimum size regulations with the regulated mesh sizes in use. Thus, for example, yellowtail flounder, with a minimum legal size of 13 inches, are caught with net meshes that retain a substantial portion of fish at and below this size (Figure 5). The example of the fate of the strong 1987 year class of Southern New England yellowtail flounder is illustrative of a number of features of the discard problem in the region. The magnitude of yellowtail flounder discards of the 1987 year class is given in Figure 6. From these discard estimates and landings-at-age data, it is determined about 77 million fish from the 1987 year class were landed through 1992, of which 46.5 million, or 60%, were discarded. Most discards occurred at ages 2 and 3, when the mesh sizes in use (4 – 5 1/2 inch diamond) retained large numbers of fish less than 13 inches in length. The foregone value of this catch due to discards is conservatively estimated to exceed $50 million.
The discard rates of sublegal sizes of species, such as yellowtail flounder and haddock, appear to vary directly with year class strength. Given low abundance of alternative ages in the stock or other valuable species, strong incoming year classes appear to attract effort, in spite of the necessity of discarding substantial fractions of the catch. Thus, one consequence of overexploited fishery systems is they encourage short-term high exploitation of incoming year classes to the ultimate detriment of long-term yields and stock biomass. In contrast, fishermen in some areas, such as the Gulf of Alaska, tend to avoid strong incoming year classes of species, such as sablefish, since the value per unit weight increases with size (R.Methot, NMFS/AFSC, personal communication). In the latter case, the presence of substantial quantities of other age groups in the exploited population offers an alternative to targeting recruits and is a direct consequence of more moderate exploitation rates.
A substantial quantity of the catch of some species is derived by recreational fishing on bluefish, summer flounder, scup, striped bass, and others. Discard rates of recreational fisheries of the region have increased in recent years owing to a combination of increased size limits, the institution of recreational bag restrictions, and a move to catch and release in recreational fisheries. Not all fish released by recreational fisheries survive, but this fraction is generally poorly understood. Understanding the potential impacts of recreational discards on mortality rates of the stocks will require a more detailed study of hook and release mortality rates.
Figure 5. Estimated values offish lenth at 50% selection (L-50), length at 25% selection (L-25), and length at 75% selection (L-75) for yellowtail flounder, based on nine research studies. Data for diamond and square mesh studies are plotted separately Selection data are plotted against the experimental cod-end mesh size used in the various studies. Predominant mesh sizes in the fishery are 4 to 5 1/2 inches, diamond
Figure 6. Estimated catch (landings and discards) of the 1987 year class of southern New England yellowtail flounder, 1988–1992. Data are presented for landings and discards separately. Total catch of the year class was 77 million fish (age 1–5), with 46.5 million (60%) discarded.
The catch of non-marketed species in many of the region's fisheries is relatively high. For example, the predominant species caught in the directed fishery for yellowtail flounder in the Southern New England region is little skate (Raja erinacea) (Murawski 1991). The species is not used for human consumption and only a limited quantity can be sold as bait. Thus, large quantities are discarded. Likewise, other trawl and gillnet fisheries of the region catch and discard large quantities of spiny dogfish (Squalus acanthias), red hake (Urophycis chuss), sculpins, sea raven (Hemitripterus americanus), four-spot flounder (Lepidorhombus boscii), ocean pout (Macrozoarces americanus), and others. Most of the discards from the Gulf of Maine sink gillnet fishery, targeted at cod, pollock and spiny dogfish, result from the catch of unmarketable species.
The catch of species not brought to market is doubly troublesome to fishermen. First, substantial quantities of fish catch are discarded, in many cases dead, for want of a market, representing some potential foregone economic opportunity. Moreover, the recent increases in abundance of some species, such as skates and dogfish, has exacerbated the problem of interference with fishing operations, since these fish are handled but generate no revenue and are becoming a larger fraction of the total catch.
The development of economic uses for these species may in some instances hold the key both for improving the efficiency of harvesting and ensuring species selective harvesting practices do not have unintended ecosystem consequences for the community of animals subject to harvesting. For those species currently underfished and uneconomic to bring ashore, more effort is needed to develop their market potential. Even if the revenues generated by such species simply cover the costs of handling, the advantages of such a strategy are bycatch species will be more fully utilized rather than “wasted” and the overall fishery economy will expand.
The trend towards more complete utilization of fishery resources has resulted in increased specialization of fisheries and harvesting techniques. Given the highly mixed nature of catches in most fisheries, an increasingly important consideration is the extent to which catch and discard patterns in one fishery influence stocks available to fisheries concentrating on the shared resources. Gulf of Maine groundfish fisheries offer an example of the magnitude of yields foregone to competitive fisheries for shared resources (Figures 7 and 8), particularly as a result of discarding. Small mesh fisheries catch and discard large quantities of small roundfish and flatfish. Gillnet and large-mesh trawl fisheries compete directly for some species. At current levels of fishing effort, the cumulative equilibrium value of groundfish landings in the region is about $90 million. What would be the value of the resource if discard mortalities were eliminated? To estimate the value of groundfish landings foregone to bycatch, the mixed-species, multi-fishery model described in Murawski et al. (1991) was re-run with discards eliminated. The relative values of landings with and without discards are described in Figure 8. The overall value of landings approximately doubles, with the majority of new benefits accruing to the two large-mesh trawl fisheries. Although only modest increases in benefits occur for the small-mesh trawl and shrimp fisheries, reductions in the discards of groundfish subsequently available to other fisheries result in substantial gains in overall benefits.
As this and other examples confirm, a holistic view of regional fisheries may offer important perspectives on the true opportunities available from bycatch mitigation. Efforts to reduce discarding of valuable flatfishes and gadoids in the shrimp fishery have recently included the mandatory use of shrimp separator devices. Other management regulations, such as increasing minimum net mesh sizes in groundfish trawl and gillnet fisheries and reductions in effective effort levels, will likely have significant overall benefits and change the distribution of benefits among the component fisheries.
Figure 7. Results of sea sampling trip reports for four Gulf of Maine groundfish fisheries in 1991. Percentages of the total catch (in weight) discarded are given by fishery. Data are presented as pounds per sampled trip discarded and kept. Source: Murawski 1993
Figure 8. Calculated ex-vessel value of six fishery units exploiting Gulf of Maine groundfish resources under current bycatch conditions and assuming no discard mortalities for groundfish species. Analyses are based on an equilibrium mixed-species fishery model described in Murawski et al. 1991. Fisheries are: Trawl OFS = offshore trawl fishery using 5 1/2" mesh, Trawl INS = inshore trawl fishery using 5 1/2" mesh, sink gillnet fishery for groundfish, a trawl fishery directed at northern shrimp, a small-mesh fishery directed at silver hake (whiting), and a small-mesh fishery directed at other species. Total benefits approximately double, assuming mitigation of bycatch.
The bycatch of protected marine mammal, turtle, and bird species in various fisheries has elicited increased scrutiny in recent years, both regionally and nationally. In general, mobile gear fisheries take few marine mammals and birds, and turtles are caught infrequently in trawl fisheries at certain times and places. Fixed-gear fisheries, however, do take these species, and the level of sampling of their bycatch and efforts to minimize the takes have increased significantly in recent years.
One of the most problematic bycatch issues concerning marine mammal bycatch is the kill of harbor porpoise in the groundfish sink gillnet fishery of the Gulf of Maine (Anon 1992a; NOAA/NMFS 1993). The fishery is pursued by about 250 small and medium-sized vessels, primarily targeting on cod, pollock, and spiny dogfish. Recent estimates indicate an annual bycatch of 900–2, 400 harbor porpoise per year, representing 2%–5% of the extant population (NOAA/NMFS 1993). Managers have proposed a goal of reducing the rate of bycatch by half, and measures to do so are currently being contemplated. Some of these potential mitigation measures involve reductions in the overall amount of effective effort in the gillnet fishery, restrictions on effort in area/time blocks where the majority of porpoise bycatch occurs, and use of acoustic beacons to deter porpoises from being captured.
Catches of dolphins, pilot whales (Globicephala melaena), and other marine mammal species, as well as turtles, have been observed and monitored in offshore fixed gear fisheries, including pelagic gillnets and longlines targeting swordfish and tunas. These fisheries result in some “takes” (either kills or non-lethal encounters with these protected resources). Some of these fisheries have been restricted, in part due to their catches of these bycatches. Sea sampling of these fisheries to document mammal and turtle bycatches has been increased under auspices of both the Marine Mammal Protection Act and the Endangered Species Act.
Comprehensive programs of at-sea sampling of catches and discards have been implemented for a number of U.S. fisheries in the Northwest Atlantic, including gillnets, pelagic driftnets, pelagic longlines, pelagic pair trawls, bottom trawls, fish and lobster pot fisheries, and dredge fisheries. The collection of adequate samples for routine discard estimation has been made for some fisheries beginning in 1989. The data derived from these sea sampling programs have allowed more complete assessments of the magnitude and causes of discards.
The bycatch and discard issues reviewed in the above section are important in many if not most of the region's fisheries. The development of methods to mitigate problematic discarding has as its basis an understanding of the processes, both ecological and technological, resulting in these mixed catches of species and size groups. Some general analyses of discard and bycatch have been undertaken with sea sampling data collected over a wide range of the region's trawl fisheries (Murawski 1993).
Otter trawl fisheries of the region account for the majority of offshore catches and target a wide variety of demersal and pelagic species. A general linear model (GLM) was used to examine factors influencing overall discard rates and other attributes based on tow-by-tow sea sampling data collected from these fisheries during 1989–1992. Data in Figure 9 represent the relative magnitude of discard rates (all species weight discarded/all species weight caught) for a wide variety of trawl fisheries sampled. The classification of target species or species groups is determined by querying the captain as to the primary species sought during the trip. The target may be either an individual species or a species group. These target fisheries span a wide range of mesh sizes, areas fished, and other fishing techniques. Interestingly, relatively low overall discard rates occur in small-mesh fisheries directed to pelagic species, such as mackerel, butterfish (Peprilus spp.), squids, and tuna. The directed fishery for skates also results in low overall discard rates. Conversely, the highest relative discard rates result from fisheries directed for flatfish. As noted previously in this chapter, high overall rates of discard from flatfish fisheries result both from the catch of non-marketed species, such as skates and others, and the catch of sublegal sized target species.
The factors influencing the species diversity of sea sampled catches, including year, month, area, primary target species, cod-end mesh, tow duration, vessel tonnage, and depth, are evaluated in Figure 10. The F statistics corresponding to variables in the GLM are presented in this figure. Species diversity is defined as the inverse of Simpson's diversity index, ranging from 0 where a single species dominates the catch to 1 where all species are equally represented. The overall tow duration is by far the most important factor in determining the diversity of the catch. These analyses support the earlier observation that the ability to target individual species and thus to minimize unwanted bycatch may be severely compromised when tow times are increased, as in response to lower catch rates in the New England trawl fishery.
The effect of cod-end trawl mesh size on the overall discard rate is presented in Figure 11. These analyses are restricted to the Southern New England-George's Bank region for cod-end meshes of 4" and greater. These data represent a wide variety of target species, areas, and specific fishing techniques. Nevertheless, the decline in discard rate with increasing mesh size is clear for these fisheries. Discards here are defined regarding both non-marketed species and undersized target species. The adoption of larger mesh sizes, as proposed for the Northeast groundfish fishery, should produce lower overall rates of discard, all things being equal.
Figure 9 Relative effect of primary species sought on total discard rates from sea sampled otter trawl trips in the Northwest Atlantic off the U.S., 1989–1992. Data plotted are the standardized coefficients from general linear models incorporating primary species sought as a categorical variable. Thus, for example, highest overall discard rates occur for the witch and yellowtail flounder fisheriess, whereas mackerel, tuna, and skate trawl fisheries have the lowest overall discard rates.
Analysis of sea sampling data from the diversity of the region's fisheries is critical to developing of effective strategies to reduce bycatch and discard interactions and determining the success of new management regulations implemented to achieve these goals. Collection of such data generally has strong support from the industry. Most sampling is now voluntary, although marine mammal and endangered species sampling can be mandated. Environmental groups also support these programs.
Figure 10 F-ratios from general linear models fitted to sea sampling data collected from otter trawl fisheries in the Northwest Atlantic off the U.S., 1989–1992. Dependent variable is species diversity (1/). N.S. = non-significant (P>0.05) factor.
Figure 11 Effects of cod-end mesh size on the proportion of trawl catch discarded for all species caught in otter trawl sea sampling in the Northwest Atlantic off the U.S., 1989–1992. Data are combined for all primary species targeted and noted in Figure 9.
The previous sections of this chapter have demonstrated high discard rates in some fisheries. A fundamental question is how these discards figure into the yield and stock potential of fisheries and the results of stock assessment calculations. Bycatch and discard data are expensive to collect, and sampling rates for discards may be substantially lower than for corresponding landings of a species, thus potentially mixing poor data with good. The case for including discards in assessments must justify the expense and effort of their collection.
The inclusion of fishery discard data in standard assessment models can in some cases drastically alter our perceptions of the status of exploitation of stocks and the balance of yields accruing from changes in regulations (Saila 1983; ICES 1986). Analytical stock assessments generally include a retrospective aspect and a prediction. Retrospective assessments often combine estimates of catch-at-age (or length) with relative indices of stock abundance from fisheries or research vessel sampling. Results of these retrospective calculations are trends in stock size and fishing mortality rates. Failure to account adequately for all components of the catch, landings and discards, may have important implications for the results (Table 30). If discards are primarily juvenile fish, then failure to account for them adequately will result in underestimates of fishing mortality, especially the stock size of the young fish. Implications for underestimating young fish discards may or may not have significant consequences for stock sizes and biomasses at older ages. Likewise, the overall goodness of fit may be improved, particularly if relative abundance indices include these young fish, but this effect may be variable. Inclusion of discards of adult fish will have a positive effect on estimates of stock biomass and to a lesser extent recruitment, stock numbers-at-age, and the overall goodness of model fit. The implications of additions of older discards for F calculations may be variable.
|AGE OF DISCARDS||F-AT-AGE||#s-AT—AGE||STOCK BIOMASS||RECRUITMENT ESTIMATE||GOODNESS OF FIT|
|+ = POSITIVE EFFECT||- = NEGATIVE EFFECT||/ = VARIABLE EFFECT|
It is important to stress discards be treated consistently in all phases of the assessment process (ICES 1986). For example, the estimation of higher recruitment levels owing to the inclusion of young fish discards would be partially compensated by higher fishing mortality rates at these young ages. The net result would produce essentially the same overall yields in some types of calculations. The importance of discards to fishery predictions depends very much on the types of predictions being made, the assumptions of discard proportions over time (constant or variable), and the exploitation patterns at age (the partial recruitment) held constant or changed. Table 31 provides a general overview of the impact of discards on various predictions under several assumptions. This table is taken from the 1986 report of the ICES Working Group on Methods of Fish Stock Assessment (ICES 1986). In the simple case of a short—term TAC forecast, assuming constant partial recruitment and a constant fraction of discards in the catch, the inclusion of small fish discards has no effect on the predictions of yield. If, however, the discards are variable from year to year, but are predictable, then discards have a moderate impact on assessment calculations. Long—term calculations, such as equilibrium yield per recruit, particularly under conditions of variable discard proportions, are the most sensitive to the inclusion of discards in the assessment. When variable recruitment is combined with models assuming a changing partial recruitment (exploitation pattern), then the results are especially sensitive to the inclusion of discards, even when the discards are a constant proportion of the catch of some age groups.
|PARTIAL RCT.||ASSESS. CALC.||CONSTANT PROPORTION||VARIABLE PROPORTION|
|CONSTANT||SHORT-TERM FORECAST||O||**||O + error|
|(e.g., CATCH FORECAST, YPR, SSB/R)|
|LONG- TERM CATCH||RELATIVE YIELD||*||***||* + error|
|FORE- CAST||RELATIVE BIOMASS||*||**||* + error|
|CHANGING||SHORT-TERM LOSSES-GAINS||**||***||** + error|
|LONG- TERM||RELATIVE YIELD||**||***||**+ error|
|LOSSES /GAINS||RELATIVE BIOMASS||*||*||* + error|
0 = NO;
* = SMALL;
** = MEDIUM;
*** = MAJOR
The use of discard estimates in routine stock assessments requires such data be available for the series of years included in the retrospective analysis. For most Northwest Atlantic fisheries, these discard estimates are available only since 1989. In some cases, direct estimates of discards via sea sampling have been supplemented with indirect methods like estimating the discarded proportions by applying commercial gear selection characteristics to population size compositions as determined from research vessel surveys. Discard estimates are routinely included in assessments of important Northwest Atlantic flatfish stocks, including yellowtail flounder, summer flounder, and American plaice (Hippoglossoides platessoides).
Figure 12 presents calculated fishing mortality rates for the 1987 year class of yellowtail flounder in southern New England. The fishing mortality on the cohort was extremely high for fully recruited ages (3+). For ages 1–3, discards accounted for a substantial fraction of fishing mortality. Effects of including or not including discards on assessment results have been examined for the stock. Failure to include discards overestimates the long-term yield and spawning potential of the stock for various assumed levels of fishing mortality and age patterns of fishing. Short-term yield forecasts are sensitive to the inclusion of discards when large year classes are present, such as the 1987 year class, but are insensitive when the stock is comprised of mostly older age groups. Similar results have been documented for other stocks. In some cases, such as silver hake, the inclusion of discard data has helped reconcile disparate trends in abundance, survival rates, and landings trends, compared to when only landings data were available.
Figure 12 Calculated fishing mortality rates at age for the 1987 year class of southern New England yellowtail flounder. Source of mortality (discards vs. landings) is subdivided based in the relative proportions of the catch in numbers. Souce: NOAA/NMFS 1993.
Bycatch, and in particular discarding, is pervasive and problematic in U.S. fisheries of the Northwest Atlantic. Discard problems are manifested in several important ways: (1) marketable species too small or otherwise prohibited from landings are nevertheless retained in the gear and are subsequently discarded, (2) species for which no current market exists are caught simultaneously with commercial or recreational species and are discarded, (3) various fleet sectors compete directly for landings or indirectly through bycatch interactions, and (4) non-fishery bycatch species including marine mammals, turtles, and birds, are inadvertently killed.
Bycatch is not a new problem faced in the region's fisheries, but the increased intensity and specialization of these fisheries and the heightened awareness of ecological implications of bycatches has placed added significance on finding approaches to their mitigation. Advantages to be achieved from successful bycatch mitigation programs include: (1) more efficient use of commercial species discarded because of size or quality, (2) establishment of markets for species not currently landed but taken as bycatch, (3) improved conservation of protected resources, and (4) increased access to resources shared by interacting fisheries.
Several regulatory approaches have been used or are being evaluated to reduce bycatch and discards. The use of a fixed grate in the northern shrimp trawl fishery has been mandatory since 1992. The intent of this technology is to allow legitimate fishing for shrimp with small mesh trawl gear, while reducing the discards of substantial quantities of undersized flatfish and roundfish. The success of this device in reducing discards is currently being evaluated through sea sampling. Another technological approach to reducing discards has been a move to larger mesh sizes in most of the region's trawl fisheries. Proposed Amendment 5 to the Northeast Multispecies (groundfish) FMP calls for 6-inch trawl mesh for most of the New England region. Furthermore, the intent of the plan is to specify minimum fish landing sizes at the 25% retention points for the 6-inch mesh, thereby resulting in virtually all of the catch being retained, it is hoped. Other gear improvements to minimize discarding are also being contemplated.
Area/time management measures have been proposed to minimize the catch of juvenile fish and to reduce harbor porpoise bycatch in gillnet fisheries. The development of a near real-time system to close areas wherein large concentrations of sublegal sized fish occur is intended to stop the practice of targeting local concentrations of incoming recruits. Such a system is heavily dependent upon current data and must be responsive enough to ensure timely closure of areas containing high concentrations of small fish.
Perhaps one of the most important measures that may mitigate high discard rates is the proposed reduction in the total magnitude of fishing effort. The Northeast Fishery Management Council has called for a 50% reduction in days at sea over a five-to-seven-year period beginning in 1994. If the reductions in days at sea translate into reduced fishing mortality rates, then the reliance on incoming year classes should decrease and be followed by a greater degree of stability in year-to-year landings.
Various methods have been proposed to reduce harbor porpoise bycatch in the sink gillnet fishery. Obviously, an across-the-board reduction in gillnet effort would have some impact, but the catch of porpoises varies temporally and seasonally. Current work in this regard has been focused on defining area/time blocks accounting for the majority of porpoise takes. By limiting effort reductions to critical times and areas, it is hoped bycatch can be reduced while minimizing adverse economic consequences to the industry. Other fisheries, including pelagic gillnets, longlines, and pelagic pair trawls, are being closely monitored for bycatch of protected species. In some cases the fisheries have been severely limited or curtailed altogether when significant catches of these species have occurred.
The discard of unmarketable species, significant in some of the region's fisheries, will remain high pending either the development of new markets for these species or the use of harvesting gear minimizing their catch.
The inclusion of discards may be important in gaining an understanding of the overall yield potentials of the region's resources and the balances of gains and losses of yields among competing fleets when new regulations are being considered. Inclusion of discards in stock assessments is considered critical in many cases, and programs for discard data collection need to be expanded. Discard sampling data are routinely being used by managers to evaluate potential changes in the management program and to monitor the success of new regulations in reducing discarding.
The productivity, diversity, and sustainability of many of the region's fisheries are currently being seriously compromised due to overexploitation and high rates of discarding. If these fisheries are to operate on a sustainable basis with higher levels of production, reductions in the effective effort and a significant shift in the exploitation pattern to older animals will be necessary. The use of some or all of the above-described measures will be pivotal in attaining these goals. Continued monitoring of these fisheries through sea sampling is necessary to evaluate the success of various management measures in reducing bycatch and discarding.
The commercial fisheries of the Northeast Atlantic, like many commercial fisheries elsewhere, have their share of bycatch problems. In the past, these have been viewed as problems interfering with the proper management of the fisheries, rather than as threats to the conservation of endangered species or to the interests of rival groups like sports fishermen. However, perceptions of the problem are changing. Public awareness of perceived bycatch threats to other species is increasing. This is particularly the case for marine mammals and birds, but is likewise true for benthos and non-target fish species. Interfleet bycatch conflicts have also intensified.
This chapter seeks to provide a brief summary of the bycatch and analogous problems in Northeast Atlantic waters. It is intended to illustrate the problems rather than to form a comprehensive review of the field. This chapter covers the following four aspects:
of the problems
The biological significance of the problems
Administrative responses to the problems
Underlying problems of stock assessment and management posed by bycatch
The examples are centered on North Sea fish stocks and European Community administrative responses because these are areas where the scientists of the region have most experience and where problems are particularly acute.
The best documented examples of commercial discarding in Europe are found in the trawl fisheries for roundfish (gadoids) in the North Sea. Here long-term at-sea sampling of discards of haddock and whiting have been carried out by the Scottish Fisheries Department. Scottish catches of these species are a sufficiently large percentage of the total international catch to allow reasonable extrapolation of these discard rates to unsampled catches. These calculations, documented in the ICES Working Group on the Assessment of Demersal Stocks in the North Sea and Skagerrak (Anon. 1993a), reveal high levels of discards in these fisheries. The discard rate has approached 50% of the total catch by weight of these species in some years (Table 32). The biological significance of this discard rate is discussed later, as are administrative initiatives to reduce it.
North Sea Haddock
|Total Catch||H.C.Landings||Discards||Industrial Bycatch|
North Sea Whiting
|Total Catch||H.C.Landings||Discards||Industrial Bycatch|
Non-target fishing mortalities are less well documented. Main areas of concern are the mortalities generated on marine mammals and birds from entanglement in gillnets and the destruction of sea floor species and habitat (killing of invertebrates) by ground trawl gears. This particular problem is associated with the heavily chained twin beam trawls used to catch flatfish, particularly Dover sole and European plaice (Pleuronectes platessa). There is also a concern with the impact of scallop dredges. These matters are often seen in the wider context of the environmental effects of fishing, encompassing not just bycatch problems, but also the perceived removal, particularly by industrial fishermen, of food species important to birds, salmon, and other species. It also covers the fisheries-generated litter problem.
Besides the problem of discard mortalities, the problem of non-catch mortality rates (unobserved fishing mortality) has also been the focus of more discussion in recent years. Biologically the problem of mortality caused by contact with gear not resulting in catches is somewhat similar to the problem of non- or wrong-reporting of catches. These all result, if not estimated, in an incomplete inventory of fishing mortality leading to wrong assessments of the state of the fish stocks. Two main reporting problems exist in European fisheries. These are black fish (non-reported catch) and grey fish (catch which is misreported as to area or species), particularly prevalent in the North Sea flatfish fisheries for a number of years. For example, higher-valued sole was being misreported as plaice to enable the quota to be overrun. While not strictly bycatch problem, under-reported catch, like the bycatch problem, may be symptoms of fishing pressure being too great.
In brief, therefore, most of the North American perceptions of the bycatch problem are also recognized in Northeast Atlantic Fisheries, but there are local variations in the types of concern, like misreporting and environmental effects. Some of these may alter the ways in which society manages fisheries and affect the public image of fishermen. The traditional European public perception of fishermen as colorful or dignified “toilers of the sea” may have begun to shift, if only slowly and slightly, towards a view of some groups of fishermen as pillagers of the environment. If encouraged to develop too far, this perception could lead to the closure of fisheries.
At-sea observer programs are comparatively rare in European fisheries. As a consequence, many of the less obvious bycatch problems tend to be unquantified. While it is possible to give the scale of the problem for some fisheries, for most it is possible to give only some indication of the existence of a problem. Table 33 attempts to set out the seriousness of bycatch and other analogous problems for the various fisheries of the North Sea, particularly acute in some fisheries, but less in others. We will therefore discuss each problem in the context of the fleet or fleets where it is of greatest concern.
In practice, splitting the North Sea fisheries into just eight fleets is a gross oversimplification, since every country, every port, and every gear type and vessel size have different mixtures of species in their catches and different size compositions of catch that depend on when, where, and with what gear configuration they fish. Rocha et al. (1991), for example, classifies the English trawl fisheries into a continuous range of catch percentages. They show trawler behavior ranges from pure fishing for flatfish through mixed flatfish and cod fisheries, fishing for cod, and to fishing for haddock, whiting, and saithe. Thus, to talk of the North Sea trawl fishery for gadoids is a considerable simplification, and the problems discussed may be more acute in some sectors of this “fleet” than they are in others.
|Fishery||Mixed Targets||Incidental Catches||Commercial Discards||Marine Mammal Kills||Bird Kills||Other Non Target Kills||Other Environment Impacts||Black Fish Landings||Grey Fish Landings|
|Trawl Fisheries for Gadoids||***||*||***||*||**||**|
|Trawl Fisheries For Flatfish||*||*||**||***||**||**|
|Industrial Trawl Fisheries for Sandeel, Pout||***||*||**|
|Industrial Trawl Fisheries for Mixed Clupeoids||***||***||**||**|
|Purse Seine Fisheries for Pelagics||*||*|
|Trawl Fisheries for Shellfish||**||***||*|
|Gill Net Fishing||*||**|
The problem of a mixture of target species is a particularly serious one for the gadoid trawl fisheries of the North Sea. The main species sought are cod, haddock, and whiting. These are fish with intrinsically different growth rates, sizes and ages of maturity. Whiting reaching maturity at age 1 and about 23 cm requires a different fishing regime than cod that is 50% mature at age 4 at a length of 60–70 cm. Since both species are caught in the same gear, the mesh size adopted has to be a compromise between the two species. This compromise must lean heavily towards the need to catch the less valuable whiting (Figure 13). Cod is thus exploited at a much too young age.
figure 13. The relative size and maturity of cod and whitting at age. Source: Pope 1993.
The alternative strategy of exploiting the cod at an older age is adopted by some groups of trawl fishermen and also gillnet and longline fishermen, but they tend to lose out to fishermen using the legal minimum mesh size of 100 mm in whiting and haddock-oriented fisheries that also catch some cod.
The mixed nature of these fisheries produces uncomfortable compromises for minimum landing sizes. These are set at 27 cm for whiting, 30 cm for haddock, and 35 cm for cod. In fact, whiting has the smallest girth to length ratio of the three species. Consequently, the tendency of fishermen to adjust the rig of their gear to catch as many minimum-sized whiting as possible inevitably leads to undersized catches of haddock and cod which must be discarded. While regrettable, this practice sometimes makes good commercial sense in the shortterm. It is, of course, nonsense in the longer term. The elimination of discards is a relatively popular aim because they are seen as a waste. Moreover, if a conservation measure could be devised that would eliminate discards without losses to the landed catch, this would produce long-term gains without short-term losses.
The desire to catch the smallest legal-size fish springs from the fact that the three species have high exploitation rates and hence the bulk of the fishable stock is close to the minimum size. A generally lower level of exploitation has been scientifically advocated for this fishery for a number of years. Among other benefits, this might help fishermen to get out of the vicious cycle of fishing for the smallest available fish and would reduce the consequent discarding of fish below the minimum size.
It is also worth considering if bycatch problems are themselves a symptom of overexploitation. Fishermen operating at low profit margins may be forced to make very short-term decisions they might avoid in more profitable circumstances. If the bank is threatening to foreclose on a vessel loan, the owner is likely to take a very short-term view of the world and not to be too concerned about threats to his longer-term livelihood.
Mixed fishery problems are also a feature of the industrial fishery for sprat. While this fishery is nominally for the small clupeoid sprat (Sprattus sprattus) it takes very large bycatches of herring (Clupea harengus). Since the herring caught are small, this has a potentially damaging effect on the directed human consumption fisheries for herring. Because these bycatches are for the most part deliberate, they are classified here as a mixed fishery problem. Alternatively, it might be viewed as a problem of incidental catch, but since the herring catch frequently exceeds the sprat catch, this would be a misleading classification.
The problem of incidental catch is more clearly seen in the North Sea industrial fisheries for demersal species, particularly in the case of the fishery for Norway pout a small (<20 cm) gadoid species found in the northern North Sea. Its distribution overlaps that of small haddock and small whiting, and bycatch mortalities are generated on the youngest ages of these species. It is a particular cause of friction with commercial fishermen who also suspect the industrial fishery for sandeel of causing similar problems for cod in the central North Sea, though available evidence does not support this.
Table 34 shows the estimates of fishing mortality by age of haddock and whiting partitioned between the mortalities caused by commercial landings, by commercial discards, and by the industrial bycatch. The scale of the discard problems and industrial incidental catch problems in these fisheries is thus well described in mortality terms.
Incidental catches also occur in trawl fisheries for shrimp and prawns (Nephrops) which take incidental catches of small gadoids and flatfish. These effects are less well quantified.
North Sea Haddock (average 1987–1991)
|Age||Total Catch||H.C.Landings||Discards||Industrial Bycatch|
As described in the previous section, commercial discards are a serious problem in the trawl fisheries for gadoids. They are also a problem in the flatfish trawl fisheries where undersized flatfish and over-quota cod are discarded. Discarding of undersized commercial species is also a feature of the trawl fisheries for shellfish. Generally these problems have not been investigated so comprehensively as in the gadoid fisheries. Recently, however, the EC has commissioned research on a number of these problems that is likely to become available soon.
Non—target mortalities of marine mammals and birds are a potential problem in the gillnet fisheries of the North Sea. Since these are mostly directed to demersal fish, such as cod, the problem is probably less acute in the North Sea than in areas like the western English Channel, Celtic Sea and Bay of Biscay where gillnets are used for pelagic species such as bass and tuna. Seabirds and harbor and grey seals (Halichoerus grypus) have generally increased in the North Sea area, suggesting mortalities caused by fishing gears are generally insufficient to cause problems at a population level. The situation with small cetaceans is much less clear. Accurate time series do not exist, but evidence points to a decline. Cetaceans are impacted by some purse seine fisheries, as well as by gillnets. Anon. (1992) summarizes the current information on all environmental effects of fishing in the North Sea area.
Another concern with the direct effects of fishing on non-target species is that of heavy ground gears on benthic species. The heavy chain ground gears used by twin-beam trawlers can penetrate some sediments to a depth of 6 cm and cause significant non-catch mortalities and discard mortalities on benthic species in the trawls' path. The global effects of such mortalities are not clear. This is because, although these gears sweep in an area greater than the total area of the North Sea each year, much of this effort may be concentrated on selected tows and areas. Daan (1991) attempts to quantify these effects by area, and Anon. (1992) gives the areas impacted by different fishing gears in the North Sea.
The non-target fishing mortality rates generated by a fleet should, in general, be no higher than the fishing mortality rates it causes on the target species. A higher mortality rate on a non-target species would imply the two species have a very similar distribution and the catchability of the gear for the non-target species is higher than for the species it was designed to catch. This seems unlikely. However, while a non-target species may suffer lower fishing mortality rates than a target species, the adverse effects of the mortality may be greater if it is impacting a species with a vulnerable life history strategy, such as a lower fecundity and a higher size and age at maturity.
These effects of fishing can be seen to some extent in the changes generated in the overall size distribution. The size composition of fish in the North Sea shows a heavy erosion of the larger sizes. Some larger species of fish have certainly become uncommon in the North Sea (common skate, halibut). Comparisons of the North Sea fish size composition with that of George's Bank can be seen in Pope et al. (1988) and show the North Sea to be much more eroded system.
Apart from direct bycatch effects, there are a number of potential environmental effects of fishing. One is the potential competition between fishermen and seabirds and other species which forage fish, such as sandeel. The lack of forage fish has produced breeding failures in bird populations on the Shetland Islands. However, it appears birds often exploit the resource at a younger age than the fishery. Hence, if fishing produces an effect, it must be through reducing the spawning stock size sufficiently to affect recruit numbers. As is the case with most fish species, the existence and form of the stock recruitment relationship for sandeel are difficult to establish, due to the large amount of variation in the recruitment series. Bailey (1991) and Bailey et al. (1991) discuss the relationship of the Shetland breeding failures to sandeel recruitment.
Another environmental effect of fishing is the positive influence of discarded fish and fish offal on the population growth of scavenging seabirds in the North Sea area. Furness (1993) calculates the total number of scavenging seabirds has increased from about 37,000 breeding pairs in 1900 to about 614,000 pairs today and attributes this increase to the availability of fish offal and discards. Positive effects resulting from discarding and non-catch mortality probably exist for benthic scavengers, but these have not been quantified.
Mixed fishery problems have been the focus of extensive biological studies in the North Sea area. The publications presented at the ICES Multispecies Symposium (Daan and Sissenwine 1991) give a good recent overview of this work. The problems can be divided into those of technical interactions and of biological interactions.
Studies of technical interactions are concerned with the fishing mortality simultaneously generated on a range of fish species and ages by the action of fishing fleets. They thus describe the practical constraints of feasible combinations of fishing mortality on different species and ages of fish. The biological interaction problem is concerned with understanding the interactions between fish species.
Technical interaction studies alone can be of value in describing the short-term effects of fisheries management. For example, the degree to which a series of fleet catch quotas are compatible with the patterns of mortality the fleets generate would be informative. Another example would be estimates of the likely short-term losses some measure, such as a mesh increase or a closed area, might cause particular fleets. The longer-term effects of management in a mixed fishery require an understanding both of the technical interactions and of the likely biological consequences of changing the exploitation on selected species and ages of fish. Such work is able to comment on the biological and economic effects of bycatches, discarded catches, and industrial bycatches.
The ICES Multispecies Working Group (ICES 1984, Pope 1991) and the Technical Measures Working Group of the Scientific and Technical Committee of the EC (STCFWG) have made considerable progress in understanding the main biological and technical interactions occurring in the North Sea. The Multispecies Working Group has collected stomach content data systematically from the chief North Sea fish predators and interpreted them through Multispecies Virtual Population Analysis (MSVPA). This is a technique using commercial fisheries catch-at-age data and fish stomach content data to estimate both the past fishing mortalities and the predation mortalities on main fish species in the North Sea. These from the basis of forward-looking simulations usable to determine the average long-term consequences of changing patterns of fishing.
The STCFWG has constructed commercial catch-at-age by specie, national fleets, quarter year, and ICES rectangles (1/2 degree of latitude by 1 degree of longitude). This enables the fishing mortality generated on a range of species by each fleet in any subarea to be estimated. These results are of considerable worth in evaluating the results of technical conservation measures, such as seasonally closed areas or areas where gear restrictions are in force.
Plans are in hand to combine the methods and results of both working groups under the recently formed ICES Long-Term Management Measures Working Group (Anon. 1993b). Some major problems remain to be solved:
Stock recruitment relationships of the various species
Biological interactions between species occurring in their early life
Migrations of fish between areas
Practical economic problems concerned with the deployment of fishing effort
This last problem is particularly important. It arises because if fishermen are faced with a new regulation, they have a number of possible options, for example, a conservation box imposed on a fishing ground so that in certain seasons only large mesh nets could be used. Such measures can be seen in the North Sea in the cod “box” and the plaice “box” on the continental coast of Europe. When such measures are imposed, fishermen have the choice of fishing in the box with the new mesh size, fishing elsewhere in the North Sea with the conventional mesh size, fishing outside the North Sea altogether, or stopping fishing. In what proportion these options will be taken up is uncertain and may well change through time. Clearly the options chosen will influence the outcome of the measure.
Certainly it will never be possible to predict exactly the outcome of regulations, but the available analyses of the North Sea fisheries do help to describe consequences and act as thinking aids. A main conclusion from the work carried out so far is that many technical measures such as closed areas have unsuspected side effects and may not achieve their primary aims to the extent intended. An example of this is the cod "box," an area in the German Bight of the North Sea designed to protect young cod in their nursery areas by allowing fishing only with a larger than standard mesh size. Preliminary calculations of its effect by the STCFWG suggested relatively little conservation benefit for cod, although the fishing mortality was reduced on younger ages of whiting, a potential predator of young cod, and on sole. Similarly, calculations of the MSWG suggest the apparent long-term conservation benefits of mesh increases in the gadoid fishery might be undermined by increases in predation caused by releasing the relatively small whiting from fishing pressures. This is a result where the existence or non-existence of a stock recruitment relationship for cod could critically affect the results.
The example given above indicates the analyses of mixed fisheries can provide counter-intuitive results. This is particularly the case for technical conservation measures where the potential exists for making dramatic shifts in the balance of exploitation of different species. A further problem with technical conservation measures is they are frequently very uneven in their impacts on different fleets. Closing an area may well cause short-term losses to fishermen who fished there previously while perhaps generating long-term gains for another group of fishermen. Problems of this sort make it difficult to maintain equality of sacrifice and so make agreement on such management measures difficult. The alternative of reducing the overall amount of fishing may be more equitable and have at least the potential to reduce fishing costs. It will not eliminate damaging bycatch problems, although it could reduce their impact.
To illustrate this latter approach, simulations of the main North Sea fisheries were made using a modification of the method of Shepherd (1988). This method uses results from multispecies virtual population analysis (MSVPA) to make steady state predictions for the various fleets and fish stocks. These take account of the known biological interactions between the species and the technical interactions between group of fleets. The fleets shown in the following illustrative model are an aggregated description of a far more complex fishery. Table 35 shows the steady state catches that might be expected from current effort levels and the catches that the model predicts following a 50% cutback in effort of the trawl fleets. These fleets are:
The human consumption fishery for cod, whiting, saithe (Pollachius pollachius), and haddock (RNDFISH-HC). This fleet uses a variety of gears, but trawls are the most important. It generates discards shown separately (RNDFISH-DI).
The small-mesh industrial fishery for fish meal conducted with bottom trawls for Norway pout and sandeel (sandlance) (INDUST-DM). This fishery also takes bycatches of whiting and haddock.
The small mesh industrial fishery for fish meal conducted with pelagic trawls for sprat (INDUST-PEL). This fishery also takes bycatches of herring.
The flatfish fishery mostly conducted with beam trawls which fishes for plaice and sole (FLATFISH F).
The other two fisheries for human consumption herring (HERRING-HC) and for mackerel (MACKEREL) are conducted mostly with pelagic gears such as purse seines. Neither of these fisheries generates bycatch, so their effort was not reduced in the model runs.
CATCH ('000 tonnes) Current Effort
CATCH ('000 tonnes) Trawl and other Demersal Effort Halved
The model is based on a series of heroic assumptions and should be viewed as illustrative of the problem rather than as a definitive description or prediction. A major difficulty in making predictions is recruitment levels have varied systematically over the past two decades and the appropriate recruitment levels to use are therefore unclear. The model uses recruitment levels which, with current effort levels, would give the average catches of the past two decades for all species except mackerel. Mackerel has been at a depressed level throughout this period and recruitment levels were set to give a catch nearer a longer-term level. Values of fish at age are used based upon recent levels.
The results of the model shown in Table 36 indicate a 50% cutback in effort for the demersal fisheries and the pelagic industrial fishery would:
Make marginal reductions in the landings of the roundfish fleet, this resulting mostly from a reduced catch of haddock, but compensated by an increased catch of saithe.
Reduce the discards from the roundfish fishery.
Reduce the catch of both industrial fisheries, also sharply reducing the by catches of whiting, haddock, and herring.
Marginally increase the catch of the human consumption herring fishery.
Leave the mackerel fishery unchanged.
Increase the yield of the flatfish fishery.
The change in the species mix reflects both the conservation benefits of the reduced effort and changes in predation levels for some species. Thus haddock and Norway pout suffer reductions in catch of more than 50%. It is a feature of biological interaction models that effort reductions do not generate as large an increase in yield as do single species models. This is because the larger stocks of predatory fish cause increases in natural mortality.
Estimates of the value of the various fisheries are shown in Table 36. These indicate effort reduction causes smaller losses in value than in catch for the roundfish fishery and correspondingly larger gains in value for the flatfish fishery, resulting from fish in both fisheries caught on average at larger and more valuable sizes. Despite the large reduction in effort, the total catch value is only marginally less than what would be obtained with current effort levels. However, the costs of four of the fleets have been halved. Table 37 attempts to show how profit levels might change with the effort change. The analysis assumes current profit levels to be 5% of the catch values of each fleet, although this is probably optimistic. With this assumption it is then possible to make a calculation of the profit that would be made if the four trawl fleets cut back their effort. Substantially increased profits are registered by the roundfish and the flatfish fisheries, but the profit remains either unchanged or only slightly higher for the industrial fleets due to increased predation on their target species and for the pelagic fleets which did not reduce their effort. The effect of halving the demersal effort would also presumably halve the mortality rates on the non-target species caught in these fisheries.
VALUE (Millions of ECU's) Current Effort
VALUE (Millions of ECU's) Trawl and other Demersal Effort Halved
|STOCK||Current Profit Est.||50% Trawl Effort Profit Est.|
* Assumes (1) current profit is 5% of catch value, (2) prices are affected by changes in fish size but not by landings volume, and (3) effort costs are reduced in line with mortality rates.
The Common Fisheries Policy (CFP) of the EC was reviewed recently after its first 10 years of operation. The official review document (Anon. 1991) is a good source of information on current perceptions of bycatch problems in the EC. Fisheries under the CFP of the EC are controlled by direct conservation measures, such as total allowable catches, and to a lesser extent by indirect limits on fishing effort. They are also controlled by technical conservation measures, such as mesh regulation, restrictions on the way gears are rigged, minimum size limits for landings, and closed areas and gear-restricted areas. A number of derogations from the main mesh size regulations are allowed, subject to restrictions on the use that can be made of the reduced mesh size. These restrictions are sometimes in the form of limitation on the minimum percentage of target species that may be caught. They may also specify a maximum limit on the percentage of the catch that can be formed of non-target species. Other derogations may be in the form of limitations on the area fished. These regulations can be seen in EC Fisheries Legislation No. 3094/86. Table 38 shows a summary of mesh derogations.
The technical conservation measures can be seen as most obviously directed towards controlling bycatch problems. Mesh size regulation is a measure to reduce bycatches and the discarding of young fish, since the choice of minimum mesh size affects the species mix caught.
|Area||Target Species||Mm||Minimum % of target species||Maximum% of protected species|
|(ICES IV and adjacent IIa south of 64°N)||All||100||100|
|West of Scotland (ICES VI)|
|(North of 56°N West of Scotland (ICES VI)||All||100||100|
|(South of 56°||All||80||100|
|Blue Whiting||40||50 (or 80 cumulative)||10|
|Deep Water Prawn|
|All Areas except south of 52°30'N|
|(1 Nov. through Feb.)||16||50||10|
|(1 March through Oct.)||-||50||10|
Restrictions on the way gears are rigged are in general designed to prevent the selective effect of the mesh size being undermined by codend attachments, such as top-side chafers. Recent legislation has discouraged the practice of ballooning codends to make them retain smaller fish. This is achieved by restrictions on the circumference of codends. Legislation has also allowed the insertion of square mesh panels with mesh sizes smaller than the minimum mesh size. These keep an area of open mesh in the vicinity of the codend and is thought to improve the selection of trawls for whiting and haddock. These panels are also used in the derogation fisheries for Nephrops (scampi) to reduce bycatches of young haddock and whiting in this prawn fishery.
Minimum landing sizes (MLS) are imposed on important commercial species to provide a simply enforced backup to mesh regulations. A number of closed or gear-restricted areas have been chosen to reduce certain bycatch problems. Herring spawning areas have been closed to avoid overfishing spawning concentrations and disturbing egg carpets. Specific areas have been chosen to protect the young ages of cod and plaice, for example, the cod and plaice boxes on the continental margins of the central and southern North Sea. These are closed to all but local vessels and to vessels using large mesh size in their nets. Similarly, the Norway pout box was created to prevent small mesh industrial fishing in areas off the east coast of Scotland where young haddock and whiting are abundant.
Derogations from the main demersal mesh regulations allow bottom trawl fisheries for some of the smaller North Sea fish species. Small mesh nets are used for catching Norway pout and sandeel, but have maximum bycatch limits on human consumption species. Recently a derogation fishery for whiting was legalized with lower limits on the target species and maximum bycatch limits on named commercial species like cod and haddock. This was designed to allow whiting to be caught in a separate fishery. In principle, this might enable mesh sizes to be increased in fisheries for cod and haddock while enabling whiting to be caught under the derogation. Unlike most other derogations, this one is backed up by a one-net rule. The one-net rule requires the catch composition to correspond to that allowed with the largest meshed net carried. This is a regulation designed to close a legal loophole existing with derogation bycatch limits, the defense against excessive catches of the bycatch species resulting from sets made with a standard mesh net fished on the same trip. While the one-net rule is not yet an EC requirement for other derogations, the U.K. has unilaterally imposed it on other national fisheries carried out under derogations.
To discourage discarding, vessels are not allowed to carry grading machines when fishing for pelagic species using small-mesh nets, but some other management measures require certain fish to be discarded. Specifically, there is a requirement to discard fish under the legal minimum landing size (MLS) and over-quota fish. In practice, fishing under derogations may also encourage discarding of non-target fish beyond the allowable percentage. The argument for legal requirements to discard is they discourage breaches of the prime management measures. For example, MLS are set to discourage undersized nets being used. Traditionally they have been set at the L25 level of the mesh size (the fish size at which 25% of the fish are retained by the mesh). Thus, on average, an MLS requires a fisherman using the minimum mesh size to discard the part of his catch his gear retains which is smaller than the level. In practice, because fishing mortalities in the North Sea are high, many of the fish may be near or below the minimum size. For some fishermen it is economically beneficial in the shortterm to try to adapt their gear to retain as many fish as possible above the MLS. This means that instead of corresponding to the L25 of the mesh, the MLS may correspond to a much higher percentage retention. This of course, implies more discards, since the gear is retaining smaller fish.
Derogation bycatch limits may also require discarding in years when the non-target species are abundant and the target species scarce. However, it has been noted arguments for increases in the bycatch percentage of a derogation are often based upon official landing statistics, so it would seem probable derogation bycatches are at best lightly policed in some countries and they may not always lead to fish being discarded.
Norway, which is not an EC member, has a discarding ban requiring undersized or over-quota fish to be landed and sold on a nonprofit basis. These regulations are backed up by periodic closures of areas where small fish are abundant. Such legislation has been discussed within the EC but the consensus seems to be that in the very mixed fisheries of the EC, such a ban would simply force discarding to continue in a hidden fashion. This would hamper the correct estimation of discarding. Moreover, the possibilities of obtaining many successful prosecutions for breaches of a discarding ban would also seem slight. Without observers on every vessel, legal defenses for discarding would certainly exist. Fishermen could claim that discarding, observed by overflight, resulted either from the gear splitting or from the catch having to be slipped for safety reasons.
A number of selective gears have been developed by European gear designers. These use various partitions and mesh sizes to produce gears able to catch target species while reducing the bycatch of non-target species. The use of such gears is certainly desirable and there is no legal bar to their use, but without rigid enforcement they will be adopted only where the fishing industry finds their use advantageous. It would be difficult legally to specify current designs of such gear for mandatory use. Moreover, by their nature the selective features of such nets could easily be negated if fishermen were unconvinced of their value or were under financial pressure to reduce short-term catch losses. Voluntary uptake by the industry thus seems the best hope for the introduction of these gears at present.
There are few regulations specifically designed to avoid non-target mortalities on marine mammals and birds. Limits on the length of gillnets are in force to reduce the bycatch of cetaceans. Experiments have been carried out with sonic reflectors on gillnets, but so far such devices are not mandatory. Closed areas have been proposed for the scientific study of the effects of trawling on the benthos and also to provide refuge for benthic species. These have yet to be adopted, although some underwater marine reserves have been set up in inshore areas.
In the section on the biological significance of mixed fishery problems, it was made clear that in order to evaluate the long-term effects of mixed fishery problems, very detailed biological and fisheries models are required. It also requires appropriate detailed datasets. This is particularly true where bycatch limiting measures, such as closed areas, need evaluation. These represent additional stock assessment tasks. They are possibly more necessary under a multinational fisheries system like the Common Fisheries Policy of EC than might be the case for a nationally managed fishery. A manager of a single-country fishery might be able to make pragmatic decisions about which regulations might help mitigate bycatch problems and to rely upon future monitoring to assess their value. In multinational fisheries, the various national administrators usually wish to understand the gains and losses to their own fleets before legislation is approved. This knowledge has obvious negotiating benefits, but equally obviously requires more sophisticated stock assessments.
Generally, the bycatch problem presents stock assessment scientists with additional difficulties. Chief of these is the need to estimate the levels of discarded fish and non-catch hidden mortalities, often difficult to do in an unbiased fashion. Moreover, since these estimates will be based upon at-sea sampling, the resulting numbers are likely to be more variable than other parts of the catch record. Including these results likely reduces the precision of the whole assessment, but leaving them out risks bias. This bias is likely to be particularly serious in the case of assessments of the effect of mesh changes.
The ICES Working Group on Methods of Fish Stock Assessment (Anon. 1986) considered the effects of under-reporting on estimates of population size and mortality made by cohort analysis and some biological reference points. They concluded from a series of simulations that constant under-reporting of catch at all ages had little effect on assessments except future catches were underestimated in line with the under-reporting rate. They also found suddenly increasing under-reporting rates led to underestimates of the current mortality rate. Under-reporting of older fish catches led to overestimated mortality rates.
In general, estimates of biological reference points were robust to these missing data. However, the change in mortality rate needed to reach the reference point was distorted in some cases because while target mortalities remained unchanged, the estimates of current mortality were altered.
No study was made of the under-reporting of young fish that would result from unestimated discards or non-catch mortalities. However, it should be expected that constant levels of such under-reporting would lead to underestimation of mortality rates on the ages affected. More abrupt increases might lead to underestimation of the mortality rate in the current year and underestimation of recent recruitment levels. Variable levels of discarding might also undermine confidence in recruitment estimates.
Mixed fisheries with high exploitation rates, high bycatch rates, and high levels of discarding are the norm in the North Sea and in a number of other areas of the Northeast Atlantic. Bycatch, more particularly discarding, is perceived as a waste of potential commercial landings. Its elimination is sometimes seen as the one conservation measure that might not involve short-term losses to fishermen, but would require the development of more selective gears than are currently available.
In the Northeast Atlantic, bycatches of non-target species, such as seabirds, seals, and small cetaceans, occur and are a cause of increased concern to the public. Detailed models of the North Sea fisheries, including the technical and biological interactions, have been developed in ICES and EC study groups. These suggest bycatch moderating measures, such as closed areas and mesh changes, may have unexpected side effects. They may benefit fish other than the species intended and may also create predation-driven changes to the system.
The controls imposed by the Common Fisheries Policy of the EC have a number of technical conservation measures designed to reduce undesirable bycatches in the fisheries. These measures include minimum mesh sizes, gear restrictions, minimum landing sizes, derogations for catching small fish, and closed and restricted areas. Some of these measures impose a legal requirement to discard fish in some circumstances.
Proper assessment of fisheries with bycatch problems requires a greater intensity of scientific investigation, particularly in the case of multinational fisheries. Discarded catches and non-catch mortalities need evaluating if the assessment are not to risk some bias.
Where bycatch discards are a problem, technical measures are often seen as the solution. These might be of two types: (1) voluntary measures adopted by fishermen when a reduction in their bycatch rate is in their own interest, for example to reduce bycatches of small fish in shrimp fisheries that are a nuisance to discard, or to avoid the capture of sensitive species such as marine mammals. This approach is most likely to be successful when the reduction in bycatch represents no economic loss to the fishermen or may even represent a reduction in costs or work. (2) Compulsory measures are more likely to be needed when the reduction in bycatch implies an economic loss to some fishermen in the shortterm. The measures that can be adopted in this case are limited to those that can be legally defined and are reasonably enforceable.
An alternative to technical approaches to resolving bycatch problems may be using direct conservation measure like reductions in capacity, effort, or catch to reduce the level of fishing mortality on target and incidental catches alike. This may be sufficient in some cases to reduce the impact of the bycatch to acceptable levels. In other cases it may increase the profitability of the fishery and serve to reduce the rate at which fishermen discount future benefits and losses, possibly increasing the time horizon of fishermen, who may then find technical measures more acceptable. Thus, this approach may increase the acceptability of measures intended to enhance long-term benefits to fishermen and the marine environment.