In 1992, 1.8 million mt of groundfish, 162,000 mt of crab, 2,500 mt of halibut, 35,000 mt of herring, and over 130 million salmon were harvested from the Bering Sea and associated waters. The diversity of resources exploited in the Bering Sea results from a variety of gear types used to target them. Trawl, longline, and pot fisheries pursue groundfish. Crab, halibut, herring, and salmon are taken in pot, line, seine, and net fisheries, respectively. Unfortunately, while each fishery/gear type focuses on the capture of one species or species assemblage, this does not preclude incidental harvest of non-target species, sexes, or size classes. Groundfish fishermen intercept crab, halibut, herring, and salmon. Crabbers encounter groundfish and take large amounts of sublegal male crabs and non-legal females. Salmon fisheries capture species of salmon other than those targeted, as well as groundfish, including halibut.
As with bycatch around the world, these non-target harvests have spawned bitter disputes between fishermen using one gear type or another. For years, such disagreements have led to regulations seeking to protect fisheries resources and the interests of specific fishing sectors.
With few commercial fisheries in existence in the North Pacific prior to the 1920s, discarding was not an important management issue at that time. Bycatch was first recongnized in the North Pacific in the 1923 United States-Canada Fisheries Halibut Convention. However, even though the Convention recognized the existence of incidental catches of halibut in other fisheries, it did little to discourage their harvest. In fact, the Convention permitted the retention of halibut for use as food for crews. If not so used, the fish were to be turned over to conservation officers when landed (Skud 1977).
Not until after the commercial halibut fishery expanded and the International Fish Commission (IFC, later renamed the International Pacific Halibut Commission) was granted the authority to regulate these fisheries by the 1930 U.S.-Canada Convention did the first of what could be referred to as bycatch regulations appear. For many years to follow, the majority of all bycatch regulations developed were created with the protection of the halibut resource in mind. Halibut and later salmon and crab took on an elite status, their protection taking precedence over developing fisheries for other species in the region.
In order to reduce the capture of small halibut in the halibut fishery, the IFC closed two halibut nursery grounds (one in Southeast Alaska and one in British Columbia) to directed halibut fishing in 1932. Shortly thereafter (in 1935), dory gear was prohibited in all halibut fisheries south of Cape Spencer, Alaska. Dories, the principal means of setting and retrieving halibut gear in the early stages of the fishery, were thought to selectively harvest small fish (Forrester et al. 1978). For this reason, a complete ban on the use of dories in all convention waters was introduced in 1944 (Myhre 1969).
In recent attempts to rationalize bycatch management in the North Pacific, various managers have suggested the possibility of permitting the retention and sale of prohibited species. In 1937, a regulation resembling the intent of this suggestion was introduced by the IFC into the setline fishery. The so-named “One-in-Seven Rule” allowed one pound of halibut to be retained and sold for every seven pounds of other fish (principally sablefish) harvested during non-halibut openings (Bell 1956). While the program appeared to be a constructive means of dealing with the halibut taken by vessels targeting sablefish, participation in the sablefish line fishery declined as the length of the halibut fishing season increased in the 1960s. In addition, the program was reportedly abused by fishermen who used the “One-in-Seven Rule” regulation to target on halibut illegally. Given these and other problems concerning costs of management and enforcement, the regulation was rescinded in 1966 (Skud 1977).
Although not introduced solely out of concern for bycatch, discards in 1938 the IFC banned the use of setnets in the harvest of halibut in all convention waters (Bell 1956). Studies in the North Atlantic had shown that setnets selected mature, spawning-sized individuals. Because some of these nets had been purchased by West Coast fishermen, the IFC wanted to avoid potential problems in the halibut fishery by restricting their use before they took a foothold in the North Pacific halibut fisheries.
While halibut management was responsible for most bycatch or bycatch-related regulations in the pre-World War II years, other fisheries requiring regulation were beginning to emerge. Trawl gear began to be permitted for use in several fisheries by the mid-1930s. Until 1935, trawls had been banned in the North Pacific, U.S., and Canada with the exception of those used to harvest shrimp. This ban was lifted in 1935 when trawls were permitted in the harvest of flounders as long as such harvest “did not capture, injure, or destroy other food fish” (Fredin 1987).
Domestic king crab fisheries begun in 1936 received their first attention from fishery managers in 1941.21 Until that time, the concept of bycatch had not existed in these fisheries, since small and softshell crab could be kept by crab fishermen. Restrictions imposed in 1941 did prohibit the retention of small, softshell, and female crabs in the commercial fisheries. Such individuals have been similarly treated in most subsequent crab regulations. A 1952 authorization to retain halibut taken in large mesh nets in Bering Sea crab fisheries was rescinded in 1961 (Knapp et al. 1983). In 1959, legal crab gear was restricted to pots (ADF&G 1991).
During the late 1940s and 1950s, a few regulations affecting bycatch were introduced in response to specific concerns, most involving halibut. These regulations prohibiting the use of nets of any kind, included trawls, to harvest halibut (Skud 1977), the retention of halibut taken by bottom trawl gear on an incidental basis (Bell 1956), the harvesting of halibut in certain areas (NPFMC 1983),22 and a series of early spring sablefish closures considered necessary to reduce incidental halibut catch (Fredin 1987). In addition, the first minimum mesh size limits were introduced in groundfish trawl fisheries in 1948 and crab trawl fisheries in 1952. In 1959, trawls were declared an illegal means of harvesting crab--presumably because of gear conflicts and the capture of sublegal sized crabs (NPFMC 1983).
Bycatch regulations as a means of fishery management became increasingly important in the 1960s as U.S. managers attempted to control the large foreign trawl fisheries in the Bering Sea and Gulf of Alaska. Previous measures, mostly confined to U.S. and Canadian fishermen, had taken up but a few pages in the regulations of the IFC (now IPHC) and U.S. Bureau of Commercial Fisheries.23 In the 1960s, bycatch regulations quickly grew in complexity as they attempted to protect the interests of U.S. and Canadian fishermen.
Most of the regulations introduced in the 1960s and 1970s were designed to protect the Pacific halibut resource or to minimize conflicts between foreign fisheries and U.S. pot and other fisheries. As early as 1959, the Japanese government voluntarily closed an area on the north side of the Alaska Peninsula to its trawl fleets in an effort to minimize the capture of crab important to U.S. fishermen in the area (Fredin 1987). In 1963, the Japanese entered the Gulf of Alaska groundfish fisheries, but restricted their vessel operations to specific areas because of U.S. concern over the effect of these fisheries' bycatch mortalities on the halibut fisheries (Fredin 1987; USFWS 1964).
21 Tanner crab regulations were first imposed in 1967.
In 1963, the first fishery observers were used in North Pacific fisheries. These observers, originating from the United States and the IPHC, were placed onboard Japanese trawlers in the Gulf of Alaska in order to monitor halibut discards (Forrester et al. 1978). Since 1963, fishery observers have played an integral role in collecting the biological information necessary to understand the level of bycatch and discards and their impacts on non-target fisheries.
Beginning in 1964, a series of bilateral agreements between the U.S. and Japan and between the U.S. and the Soviets became the basis for fishing and bycatch management in the North Pacific. Time and area bans, as well as gear prohibitions, increasingly restricted distant water operations in the region. Milestones included the elimination of the tanglenet fisheries in 1973, the first regulations associated with marine mammal bycatch, and the imposition of a pot sanctuary and late winter and early spring trawl closures in the Bering Sea.
Bycatch management became even more volatile in the region after the passage of the Fishery Conservation and Management Act (FCMA) by the U.S. Congress in 1976. Bycatch management prior to passage of the Act was primarily a function of negotiating catch and bycatch concessions from foreign fishermen. Following passage of the Act, as the catch and bycatch needs of U.S. fishermen increased, many of the protective regulations designed to control bycatch levels in foreign fisheries were simply transferred from foreign interests to developing U.S. groundfisheries. Groundfish harvests by U.S. interests grew from almost zero following passage of the FCMA to close to two million mt a decade and a half later. Meanwhile, bycatch regulations progressed rapidly from simple instructions to try to avoid or minimize the capture of selected species, to caps limiting the total removal of prohibited species, eventually to the present system combining caps with time/area closures and various other management controls. The process moved quickly from the reassignment of catch and bycatch from distant water fleets to U.S. fleets and ultimately led to the heated debates between U.S. fishermen observed today.
The North Pacific Fishery Management Council (NPFMC), the entity responsible for managing the groundfish resources of the U.S. in the Northeast Pacific region, which shares authority with the Alaska Department of Fish and Game for crab issues and the International Pacific Halibut Commission for halibut, has been embroiled in recent years in the disputes involving bycatch issues. A series of groundfish management plan amendments24 limited the takes of crab, halibut, salmon, and herring (Clupea pallasi) in the groundfish fisheries and allocated these incidental harvest quotas among gear types in the groundfish fishery. Mortality
Throughout the 1980s, as bycatch regulations were developed, revised, and reissued for domestic fishermen in the Bering Sea, one fact continued to be recognized-better discard catch data were needed. In order to improve fishery management, protect the stocks, and ensure the commercial viability of all fisheries in the region, information concerning bycatch in each of the fisheries was considered particularly relevant. During the transitional period in the mid-1980s when U.S. fishermen began to replace the large-scale distant water fisheries of foreign nationals which had 100% observer coverage, only spotty target and non-target catch data were recorded for the growing U.S. fleet. Reduced observer coverage made it difficult for management entities to evaluate impacts of non-target mortalities, be they on fish or fisheries.
This situation changed in 1990 when observer coverage was mandated on groundfish vessels operating in the Bering Sea and Gulf of Alaska. Vessels greater than 125' overall were required to carry a biological observer 100% of the time. Thirty percent of the effort of boats smaller than 125' and larger than 60' was to be monitored. Shoreside processing plants were also faced with observer coverage requirements of varying levels. Observers were required to have academic backgrounds in fisheries or related biological sciences and underwent training courses conducted by the U.S. National Marine Fisheries Service. Potential observers were hired by independent observer contractors. Individual vessels or fishing companies contracted with the companies for observers and paid them directly at rates ranging from $3,500 to $5,000 per month per observer.
The observer program addressed many of the information concerns prevailing before the program began. Large amounts of data on target and non-target catch volumes in the target fishery have been collected. Unit weights and lengths, sexual maturity, and other data necessary for determining the biological characteristics of individuals taken and discarded in the fishery are now available. Important discard data are accessible in near-real time. Consequently, the resulting information can be used for short-term fishery management decisions, a capability of ever-increasing importance, given the overcapitalized and Olympic nature of the Alaskan groundfish fisheries. Such decisions include the identification of bycatch “hot-spots” for vessels to avoid, bycatch quota management on a week-to-week basis, and vessel-specific incentive/enforcement plans.
Of course, the data of greatest relevance to our interests in this study are those offering insights into the bycatch volumes and rates of discards occurring in the region. Although many of the gear types in operation around the world are deployed in the Bering Sea, the extensive datasets available for the region may shed light on the potential level of discard rates associated with similar fishery/gear type combinations in existence in other fishing regions of the globe.25
At the time of this writing, the most current full year for which data are available from the Bering Sea observer program is 1992. Table 21 documents total retained and discarded weight and number for four gear types and nine target fisheries in the Bering Sea. Discard rates on a weight per weight and number per number basis are also presented. Discard values include all species, sizes, and sexes, and so-called “prohibited species,” such as salmon, herring, and halibut. Likewise, retained catch includes all landed target and non-target species.
| Gear | Target | Discarded Weight (mt) | Discarded Number | Retained Weight (mt) | Retained Number | Discard Weight per Weight | Discard Number per Number |
|---|---|---|---|---|---|---|---|
| Line | Cod | 19,700 | 20,141,857 | 90,273 | 26,656,727 | 0.22 | 0.76 |
| Sablefish | 819 | 1,540,070 | 2,119 | 689,297 | 0.39 | 2.23 | |
| Turbot | 26 | 49,961 | 73 | 25,691 | 0.36 | 1.94 | |
| Jig | Cod | 7 | 17,336 | 1 | 192 | 9.85 | 90.29 |
| Pot | Cod | 215 | 702,593 | 13,205 | 3,838,416 | 0.02 | 0.18 |
| Sablefish | 1 | 2,312 | 1 | 146 | 2.20 | 15.84 | |
| Trawl | Atka mackerel | 6,691 | 13,998,929 | 42,342 | 53,925,476 | 0.16 | 0.26 |
| **>20% Pollock | 29,300 | 111,367,796 | 138,882 | 170,243,467 | 0.21 | 0.65 | |
| Cod | 23,779 | 80,874,227 | 42,189 | 17,971,591 | 0.56 | 4.50 | |
| Rockfish | 3,774 | 10,026,051 | 12,574 | 20,494,278 | 0.30 | 0.49 | |
| **>95% pollock | 40,651 | 120,243,881 | 1,043,800 | 1,318,636,387 | 0.04 | 0.09 | |
| Rock sole | 29,481 | 201,256,455 | 20,571 | 47,314,112 | 1.43 | 4.25 | |
| Sablefish | 2 | 4,614 | 28 | 8,150 | 0.07 | 0.57 | |
| Yellowfin sole | 56,638 | 353,797,480 | 101,001 | 304,743,648 | 0.56 | 1.16 | |
| Totals | 211,084 | 914,023,562 | 1,507,058* | 1,964,547,578 | 0.14 | 0,47 |
Important contributors to the regional total discard weight include the cod, pollock, and yellowfin sole trawl fisheries accounting for 11%, 33%, and 27% of regional bycatch weights, respectively. These same three fisheries accounted for 71% and 64% of total discard weight and numbers in 1992.
On a weight basis, the cod jig, sablefish pot, and rock sole trawl fisheries all discard more tonnage and numbers of fish than they retain, but the cod jig and sablefish pot fisheries are very small and contribute little to total regional discards. The rock sole fishery accounts for 15% of total Bering Sea discard weight and 23% of discard number, even though the fishery itself lands only a little over 1% of the regional groundfish total. Discard numbers from the sablefish line, turbot (Atheresthes stomias) line, cod trawl, and yellowfin sole (Limanda aspera) trawl also exceed retained numbers. Discarded weight for these fisheries is considerably less than retained weight, since their bycatch consists of high numbers of small individuals.
Low weight-based discard rates were identified for the pot cod (2%), >95% pollock trawl (4%), and sablefish trawl (7%) fisheries. The >95% pollock trawl operation is a major fishery responsible for more than two-thirds of all groundfish landings. Nevertheless, its low bycatch rate produced total discards only 19% of the regional discard weight and 13% of the regional discard number in 1992. Conversely, the >20% pollock trawl classification, although responsible for only 12% of total pollock harvests, accounts for nearly 42% of all discards generated by the two pollock categories. NMFS data can also differentiate between on-bottom and pelagic fisheries for pollock. When this is done, the quantity of non-pollock discarded from the pelagic fishery represents just 0.06% of the total catch.26
The discards recorded for each fishery include sublegal or undersized target species, as well as non-target individuals. In the pollock categories, for example, pollock alone account for about 60% to 70% of total discards. Sizable amounts of discards of Pacific cod, rock sole, arrowtooth flounder, and other flounders also occur in the pollock fishery. When these and other species are added to the mix, total discards in the pollock grouping rise to roughly 70,000 mt in conjunction with the 1.18 million mt retained harvest of pollock.
Like the pollock fishery, Atka mackerel (Pleurogrammus monopterygius) discards represent approximately 60% of total discards in the Atka mackerel fishery. The situation is, however, quite different for harvests of other species by trawl gears and harvests by other gear types. Discards of small cod in the cod trawl fishery totaled 2,165 mt in 1992. If discards of pollock, rock sole, arrowtooth flounder, Atka mackerel, and other species are added to this total, discards mount to 23,779 mt, more than ten times higher than cod discards alone. Yellowfin sole, rockfish, and rock sole discards account for 44%, 20%, and 32% of total discards in their respective trawl fisheries. In the line fisheries, cod account for 1% of total cod line discards while sablefish are responsible for 0.1% of all sablefish line discards. Halibut is a key contributor, comprising approximately 30%, to the cod line discards. In the sablefish fishery, various flounders and skates account for more than 60% of the non-sablefish discards.
26 Discards of small pollock are not delineated with this breakout.
Several species are harvested by more than a single gear type in the Bering Sea. In these cases it is interesting to look at how discard rates vary across the targets. Cod taken in line, pot, jig, and trawl operations offers a good basis of comparison. Discard rates (by weight) associated with the harvest of cod range from 9.85 for the tiny cod jig fishery to 0.56 in the trawl fishery, 0.21 in the line fishery, and 0.02 in cod pot operations. Thus, more than two orders of magnitude difference in discard rates occur between gear types targeting on the same species.
Similar variations can be observed in discard rates across the range of other species harvested and gear types used in the Bering Sea. Table 21 demonstrates that no single major gear type in use in the region has consistently high or low discard rates. Trawl rates range from 0.04 to 1.43, with an average of 0.41 kg discarded per kg retained. The weighted mean discard rate (weighted by the retained tonnage of each trawl fishery) for trawl operations in the Bering Sea drops to 0.15 kg/kg owing to the high retained tonnage and low discard weight in the pelagic pollock trawl fishery. Longline rates show the least variance, rising from a low of 0.22 kg/kg in the cod line fishery to a high of 0.39 kg/kg in the sablefish fishery. The weighted mean discard rate for line gear is 0.22 kg/kg, somewhat higher than in the trawl fishery. Pot fisheries, meanwhile, show the greatest range in discard rates, varying by more than two orders of magnitude between the low cod and high sablefish pot discard levels.
The quantities of each target species retained and discarded in 1992 by all trawl and line gears in the Bering Sea are provided in Tables 22 and 23. The tables include discards of prohibited species of halibut and herring. For the Bering Sea trawl fisheries, the retained catch weight amounted to 1,341,440 mt and the discarded catch weight 204,491 mt. Additional catches in numbers are shown for crab and salmon.
| Species | Retained Catch(mt) | Discarded Catch(mt) | Discarded/Retained Ratio(%) |
|---|---|---|---|
| Pollock | 1,115,730 | 75,734 | 0.07 |
| Sablefish | 70 | 5 | 0.07 |
| Atka mackerel | 39,670 | 6,453 | 0.16 |
| Pacific cod | 57,977 | 11,265 | 0.19 |
| Pacific Ocean perch | 10,559 | 2,218 | 0.21 |
| Yellowfin sole | 90,885 | 31,252 | 0.31 |
| Flounder | 5,044 | 24,007 | 0.48 |
| Turbot | 317 | 223 | 0.70 |
| Rocksole | 18,591 | 21,829 | 117.00 |
| Other | 1,980 | 14,664 | 741.00 |
| Rockfish | 78 | 631 | 810.00 |
| Arrowtooth flounder | 539 | 6,738 | 1,250.00 |
| Halibut | 0 | 5,136 | |
| TOTAL GROUNDFISH | 1,341,440 | 204,491 | 0.15 |
| OTHER | |||
| Herring | 0 | 4,386 | |
| Salmon (number) | 0 | 3,695,940 | |
| Tanner crab (number) | 0 | 145,344 | |
| King crab(number) | 0 | 38,108 |
Discard rates for Bering Sea crab fisheries observed during 1991 and 1992 are provided in Table 24. The number of individuals discarded per retained male is relatively high for king crab, exceeding 7.2 per retained male for golden king crab caught near Dutch Harbor and nearly six discards per retained male in the Bristol Bay red king crab fishery in 1992. Reeves (1993) estimates that in 1992 over 16,000,000 red king crab were discarded in the crab pot fisheries.
| Species | Retained Catch(mt) | Discarded Catch (mt) | Discard/Retained Ratio(%) |
|---|---|---|---|
| Pacific cod | 89,607 | 283 | 0.01 |
| Sablefish | 1,885 | 20 | 0.01 |
| Rockfish | 295 | 61 | 0.21 |
| Yellowfin sole | 0 | 67 | 100.00 |
| Atka mackerel | 19 | 23 | 121.00 |
| Turbot | 294 | 780 | 265.00 |
| Rocksole | 9 | 323 | 356.00 |
| Other | 938 | 8,131 | 867.00 |
| Pacific Ocean perch | 7 | 81 | 1,157.00 |
| Flounder | 244 | 2,030.00 | |
| Arrowtooth flounder | 1,336 | 2,470.00 | |
| Pollock | 73 | 2,558 | 3,504.00 |
| Halibut | 0 | 6,958 | |
| TOTAL GROUNDFISH | 93,190 | 20,572 | 0.22 |
| OTHER | |||
| Salmon (number) | 0 | 148 | |
| Tanner crab (number) | 0 | 18,955 | |
| King crab (number) | 0 | 2,425 |
| Year | Species/Fishery | Number Discarded per Number Retained |
|---|---|---|
| 1990 | Adak golden king crab | 5.126 |
| 1990 | Dutch Harbor golden king crab | 9.948 |
| 1991 | Dutch Harbor golden king crab | 7.23 |
| 1991 | Bering Sea snow crab | 0.106 |
| 1991 | Bering Sea Tanner crab | 0.895 |
| 1991 | Bristol Bay red king crab | 1.583 |
| 1991 | St.Matthew blue king crab | 2.036 |
| 1992 | Bristol Bay red king crab | 5.889 |
A closer look at some of the specific fisheries generating bycatch within the Bering Sea presents some interesting results. If discard rates recorded in the 1992 pollock fishery had held for earlier years,27 pollock discards in the pollock fishery could have ranged from 88,000 to 130,000 mt and from 140 million to more than 350 million individuals annually over the last 15 years. During this period, the pollock fishery in the Eastern Bering Sea has yielded annual harvests in excess of 1 million mt since 1984 and greater than 900,000 mt since 1970. Particularly large harvests in excess of 1.5 million mt occurred in the early 1970s with the recent peak of 1.35 million mt taking place in 1990. In the face of these discard and landing values, stock biomass rose from 4.3 million mt in 1979 to 9 million mt in 1985 before stabilizing in the range of 6 to 8 million mt in recent years (Quinn and Collie 1990).
Although pollock discards have reached as high as 350 million individuals and 130,000 mt during the last 15 years, discards, on average, represent less than 1% of the exploitable pollock biomass in most years.28 Although the numbers of pollock discards are high, the effort is negligible when compared with mortalities from other sources.
The estimated 1992 mortalities resulting from discarding major commercial target species in the Bering Sea is provided in Table 25. The annual mortality attributed to pollock discards is 0.016,29 about 10% of the annual mortality due to landings. The increase in total instantaneous mortality (z) resulting from discards is less than .01. Discard-induced mortalities for cod, Atka mackerel, sablefish, and Pacific Ocean perch are also relatively small in relation to mortalities resulting from landings of these species, and the increase in total instantaneous mortality is substantially less than one for each of these species.
28A high of 1.6% was reached in 1992.
29Assumes 100% mortality for discards.
| Species | Mortality Due to Discards | Mortality Due to Landings |
|---|---|---|
| Pollock | 0.016 | 0.171 |
| Pacific cod | 0.013 | 0.190 |
| Atka mackerel | 0.008 | 0.053 |
| Rockfish | 0.004 | 0.008 |
| Yellowfin sole | 0.012 | 0.046 |
| Sablefish | 0.001 | 0.053 |
| Greenland turbot/ Arrowtooth flounder | 0.012 | 0.014 |
| Rock sole | 0.015 | 0.027 |
| Flounder | 0.020 | 0.024 |
| POP | 0.005 | 0.035 |
In several cases, however, discard mortalities are sizable compared to landings mortalities. Annual mortality due to discards of turbot/arrowtooth flounder and the flounder group nearly matches landings. The rock sole and rockfish discard mortality rate is one-half of the respective landings mortality rate. The yellowfin sole annual discard mortality rate of 0.012 is one-quarter of the annual landings mortality figure. However, for each of the above species, the combined fishery discard and landing mortality rates are far below what are required to achieve the authorized Allowable Biological Catch (ABC).
Managers use a variety of techniques to adjust for the consequences of discard mortalities, and it is noted that for several species in the Bering Sea and Gulf of Alaska the losses due to discards are deducted from authorized catches. It is also apparent in the development of resource management plans that when such losses are known they are accounted for in relevant authorized catches. In the Northeast Pacific, the IPHC-approved method for adjusting the halibut stock for bycatch losses (discard mortalities) is to reduce harvest in the directed line fishery “such that the reproductive potential of the exploitable component of the stock would be the same after bycatch as it would have been if bycatch had not occurred” (Sullivan et al. 1993). According to Trumble et al. (1993), if reproductive compensation is done correctly, and if bycatch mortality is estimated correctly the halibut spawning stock should remain the same. The directed fishery pays for maintenance of the catch through the reduced quota.
Unlike many of the groundfish fisheries, the king crab fishery is an instance where discards may have done much to shape current stock conditions. King crab discard numbers in the crab fisheries as a percentage of estimated population numbers range from 20% to 40% in most years since 1980 and approach 80% in some years. Even at a low mortality rate of 25% for the discarded sublegal and female individuals, these discard levels could confront the king crab stocks in the region with substantial additional mortalities (Reeves 1993)--and this is without considering the losses associated with crab bycatch in the trawl fisheries (Table 26). It is perhaps not surprising king crab fisheries have struggled to reach 10 million lb of landings since their 1981 collapse from annual harvests in excess of 100 million lb.
| Year | Population (number) | Discard Percent (number to number) | Discard Mortality Rate |
|---|---|---|---|
| 1980 | 229,300,000 | 82.97% | 41.49% |
| 1981 | 189,900,000 | 32.34% | 16.17% |
| 1982 | 261,300,000 | 4.61% | 2.31% |
| 1983 | 89,200,000 | 4.10% | 2.05% |
| 1984 | 172,700,000 | 2.04% | 1.02% |
| 1985 | 39,900,000 | 9.57% | 4.79% |
| 1986 | 39,900,000 | 22.10% | 11.05% |
| 1987 | 75,700,000 | 11.78% | 5.89% |
| 1988 | 39,700,000 | 13.91% | 6.96% |
| 1989 | 51,100,000 | 15.95% | 7.98% |
| 1990 | 52,200,000 | 32.70% | 16.35% |
| 1991 | 43,900,000 | 39.16% | 19.58% |
Using recent data collected in the on-board crab observer program in the Bering Sea, Reeves (1993) confirms earlier work by Alverson (1980) which had identified the benefits of smaller minimum size limits in the king crab fishery. Lower minimum sizes would permit retention of many individuals which might die if discarded. It is important to note that there are no solid data on discard mortality levels and the levels used in Table 26 cannot be verified. Regardless, catch quotas represent an important but incomplete fraction of crab mortality in the crab fisheries.
Tables 22 and 23 provide data on the catch and discard of salmon from the bottom fisheries in the Bering Sea region. Erickson and Pikitch (in press) have also studied the capture and discard of chinook salmon in fisheries along the West Coast of the U.S. (California, Oregon and Washington). They report a chinook salmon catch of 7,761 for 1987, or approximately 1.4% of those years' commercial ocean landings (troll gear). Although salmon bycatch has been reasonably well documented for trawl fisheries (NRC 1990), it is perhaps less evident that seine, hook and line, and gillnet salmon fisheries have their own bycatch problems.
Other than data associated with trawling, NRC was unable to locate bycatch discard data for the Bering Sea salmon fisheries. In waters to the south of the Bering Sea, the incidental catch of juvenile chinook salmon in the Puget Sound purse seine fishery (Area 7/7A) between 1976 and 1985 varied from 111% to 489% of the adult chinook catch in this fishery, but the number of juveniles as a fraction of all target salmon species catch in numbers was much lower (0.57% to 4.39%). The situation in Puget Sound areas 8–13 regarding juvenile discards was much worse, ranging from 434% to 2867% of adult salmon catch (Shepard 1994). Riddel and Fraidenburg (1987) estimate the unobserved bycatch of chinook salmon, as a function of legal harvest numbers taken in West Coast and Alaska salmon purse seine fisheries from 1977 to 1985 (Table 27):
| Southeast Alaska | 155.3% |
| Northern British Columbia | 75.4% to 135% |
| Vancouver Island | 292% to 411% |
| Georgia Strait | 19.1% to 107.4% |
| Puget Sound | 250% to 290.1% |
Again, these numbers constitute bycatch as a function of the chinook legal harvest, not of the overall salmon catch which is much lower.
| AREA | PERIOD | SUBLEGAL | LEGAL RELEASED | ESTIMATED UNOBSERVED BYCATCH | TOTAL INCIDENTL CATCH NUMBERS | PERCENTAGE OF LEGAL HARVEST NUMBERS |
|---|---|---|---|---|---|---|
| Southeast Alaska | 1977–1982 | NA | 18,206 | 10,100 | 28,306 | 155.30% |
| Northern British Columbia | 1977–1982 | NA | NA | 33,100 | 33,100 | 75.4% |
| 1983/1984 | NA | NA | 22,400 | 22,400 | 100.00% | |
| 1985/1986 | NA | NA | 36,700 | 36,700 | 135.67% | |
| Vancouver Island | 1977–1982 | NA | NA | 31,300 | 31,300 | 292.52% |
| 1983/1984 | NA | NA | 11,100 | 11,100 | 411.11% | |
| 1985/1986 | NA | NA | 34,085 | 34,085 | 261.79% | |
| Georgia Strait | 1977–1982 | NA | NA | 16,700 | 16,700 | 52.85% |
| 1983/1984 | NA | NA | 21,900 | 21,900 | 107.35% | |
| 1985/1986 | NA | NA | 4,800 | 4,800 | 19.05% | |
| Puget Sound | 1977–1982 | 146,480 | NA | NA | 146,480 | 290.06% |
| 1983/1984 | 102,520 | NA | NA | 102,520 | 250.05% | |
| 1985/1986 | 101,510 | NA | NA | 101,510 | 285.94% | |
For troll salmon fisheries, Riddel and Fraidenburg (1987) report estimates of unobserved bycatch of 25% to 69.2% of legal harvest numbers, depending on area and year observed. Most of the discarded salmon were sublegals. Sports fisherman records were not much better and produced an estimate of incidental mortality associated with legal harvest numbers of 29.6% to 102.6% in British Columbia waters.
Although the bycatch of chinook salmon is obviously relatively low when considering the take of all salmon by these fisheries, the numbers nevertheless demonstrate the severity of the problem confronting managers dealing with this mixed species fishery. We have noted only some of the observed rates of bycatch for chinook, involving only species of salmon, but similar efforts may occur in other minor stocks, such as coho and steelhead, in particular river systems Furthermore, the bycatch of other fish species and birds is a problem in some areas. The gillnet catches of steelhead on the Skiena River, British Columbia, Canada, are the basis for a confrontation between sport and commercial fishermen. Similarly, harvesting pink salmon (Oncorhynchus gorbuscha) can in some locations have harmful effects on coho, chum (Oncorhynchus keta), and sockeye (Oncorhynchus nerka) salmon fisheries. In many of the above cases, however, the problem is dealing with mixed stocks as a bycatch issue, but often not as a discard issue (Larkin, pers. comm.). The extent of impacts and management conflicts resulting from bycatch in salmon fisheries in the Bering Sea are unknown, but given what has been observed in other areas is undoubtedly a part of the pervasive nature of the overall bycatch issue.
In most Bering Sea fisheries, the bulk of the catch is comprised of a relatively low number of species. Nevertheless, it is also true that in most of the fisheries, the dominant species are frequently accompanied by dozens of other species. While contributing little to catch weight or numbers, these other species expand the diversity of the harvest tremendously and provide considerable data concerning the overall species assemblage of the region.
Table 28 documents the number of different species taken in the various target fisheries in the Bering Sea during 1992 and presents Simpson's diversity index (Elseth and Baumgardner 1981) and Caddy's (1989) index of gear selectivity. The number of species taken in line, pot, pelagic trawl, and bottom trawl fisheries ranges from a low of twelve in the Atka mackerel pelagic trawl and arrowtooth flounder pot fisheries to a high of 119 in the pollock pelagic trawl fishery. On a gear-specific basis, pot gear intercepts the lowest mean number of species (29) but also exhibits the greatest amount of variation around that mean (SE = 15, n = 3). Conversely, non-pelagic trawl gear takes the highest mean number of species (81), but the data used to compute this mean show lower variability than for any other gear type. With the exception of the turbot bottom trawl fishery, the number of species taken by non-pelagic trawl ranges from a low of 72 to a high of 112. This range is much tighter than the spread from 34 to 108 species noted for line gears and the 12 to 119 noted for the pelagic trawls.
| Fishery | Gear Type | Number of Species | Caddy's Index | Simpson's Diversity Index |
|---|---|---|---|---|
| Rockfish | Line | 34 | 0.029 | 0.909 |
| Sablefish | Line | 36 | 0.028 | 0.702 |
| Arrowtooth flounder | Line | 38 | 0.026 | 0.807 |
| Turbot | Line | 42 | 0.024 | 0.568 |
| Pacific cod | Line | 108 | 0.009 | 0.352 |
| Turbot | Non-pelagic trawl | 26 | 0.038 | 0.539 |
| Arrowtooth flounder | Non-pelagic trawl | 72 | 0.014 | 0.821 |
| Atka mackerel | Non-pelagic trawl | 75 | 0.013 | 0.211 |
| Rock sole | Non-pelagic trawl | 76 | 0.013 | 0.645 |
| Rockfish | Non-pelagic trawl | 84 | 0.012 | 0.643 |
| Pollock | Non-pelagic trawl | 102 | 0.010 | 0.458 |
| Pacific cod | Non-pelagic trawl | 104 | 0.010 | 0.456 |
| Flatfish | Non-pelagic trawl | 112 | 0.009 | 0.537 |
| Atka mackerel | Pelagic trawl | 12 | 0.083 | 0.251 |
| Rockfish | Pelagic trawl | 14 | 0.071 | 0.677 |
| Rock sole | Pelagic trawl | 30 | 0.033 | 0.691 |
| Arrowtooth flounder | Pelagic trawl | 34 | 0.029 | 0.699 |
| Flatfish | Pelagic trawl | 35 | 0.029 | 0.793 |
| Pacific cod | Pelagic trawl | 38 | 0.026 | 0.270 |
| Pollock | Pelagic trawl | 119 | 0.008 | 0.022 |
| Arrowtooth flounder | Pot | 12 | 0.083 | 0.756 |
| Turbot | Pot | 16 | 0.063 | 0.364 |
| Pacific cod | Pot | 59 | 0.017 | 0.116 |
| Gear Type | # of Species Mean | Caddy Index Mean | Simpson Index Mean | |
| Line | 51.6 | 0.023 | 0.668 | |
| Non-pelagic trawl | 81.4 | 0.015 | 0.539 | |
| Pelagic trawl | 40.3 | 0.066 | 0.486 | |
| Pot | 29.0 | 0.054 | 0.412 |
The high number of species intercepted by bottom trawls is not surprising. Bottom trawls at times capture species living in the sediments, fixed to the sediments (for example, corals, sponges), and epibenthic forms, as well as a variety of near-bottom fishes and invertebrates. By comparison, pelagic trawls, when fished off bottom, generally avoid contact with most bottom or near-bottom organisms. The exception seems to occur when targeting pollock with pelagic gear, perhaps indicating occasional net contact with the bottom.
Caddy derived an index of gear selectivity (I = 1/S where S is the number of species taken by a gear type) to investigate the selectivity characteristics of fisheries operating in the British Virgin Islands (BVI) (Caddy 1989). His values ranged from a low of 0.03 in the fish trap fishery to a high of 1 in the jig fishery. Values approaching 1 indicated increasing gear selectivity, that is, if a gear took only one species, its index value would equal 1, whereas if it took ten species, its index would equal 0.10. Because the indices were derived only from landings of commercially important species and ignore discarded species, Caddy suggests the values observed overestimate the true selectivity of the gear.
Data from the Bering Sea fishery support Caddy's hypothesis. Bering Sea values reach a high of 0.083 for a pot fishery, compared to a BVI jig fishery with its value of 1, and fall to a low of 0.008 and 0.009 in the pollock pelagic trawl and flatfish bottom trawl fishery, respectively. Most BVI values calculated by Caddy are much higher than Bering Sea figures.
Longline and pot gear were used in both areas. Bering Sea longline fisheries intercept from 34 to 108 species, for a mean selectivity of 0.023, compared to the seven species, with a selectivity of 0.14, landed by longline operations in the BVI. An interesting exception to the rule of lower Bering Sea and higher BVI selectivity indices involves the pot/trap fisheries. The BVI fish trap value is actually lower than most of the index values for Bering Sea fish pot fisheries. This probably reflects there is a greater species diversity in the tropical fishery or perhaps a large number of the fish taken in BVI traps are landed and thus subject to study.
Counts of the number of species taken in a gear provide valuable information on the species selectivity of one gear type compared to another. But species counts say nothing about the relative abundance of each species in the catch. For example, if a gear type catches twenty incidental species, but catches only one individual of each species, does this mean it is more selective then a gear type that takes only one additional species, but at a number equal to the non-target species? Use of Simpson's diversity index (C=1-∑p2, where pi is the proportion of each species in the catch) provides a different measure of the diversity characteristics for the Bering Sea fisheries. If many species of equal abundance are taken in a fishery, the diversity index within that fishery will approach 1. As fewer species are taken or if a small number of species account for a high proportion of the take, then diversity will fall.
Simpson's diversity index measurements for the Bering Sea fisheries range from 0.02 in the pollock pelagic trawl to 0.91 in the rockfish line fishery. Although the pelagic trawl fisheries take and discard a large number of species, the overall catch of the target species is very high while the quantity of discards is very low. In the trawl rockfish fishery, several species of rockfish are targeted. This strategy lends itself to a more even distribution of the catch among different species and a higher diversity value than for fisheries where a single species is targeted.
Results of the diversity analysis suggest conclusions about species diversity by gear types can be better understood using a combination of the Caddy and Simpson diversity indices. In the Bering Sea, pot fisheries intercepted, on average, far fewer species than did the bottom trawlfisheries (29 vs. 81). Nevertheless, the Simpson diversity index for the bottom trawl fishery is only slightly above that for the pot fishery and is actually lower than the value derived for the line fisheries. Thus, while bycatch and discards in the bottom trawl fishery may affect a broader component of species in the Bering Sea than pot and line gears, many of the species discarded from the trawl fishery account for only a very small fraction of the total discards and thus have little impact on the diversity index value. Conversely, in the line fisheries the catch is more evenly distributed across species and diversity values rise accordingly. This occurs despite the fact the numbers of species taken in the line gear are lower than in the non-pelagic trawl.
In comparing the numbers of species/families of fishes taken and the distribution of mass among species for the Bering Sea trawl fishery and similar data from tropical shrimp trawl fisheries, it becomes readily apparent that distribution of the catch (by weight) for bycatch species in the tropical shrimp fisheries spreads out over many more species. In the Bering Sea pollock fishery, 14 species comprised 98.9% of the discard catch (by weight). By comparison, in the Australian northern prawn fishery and the Gulf of Mexico shrimp fisheries 14 families (many species) comprised only 86.5% of the discard catch (Table 29).
Table 29. Cumulative percent of top fourteen bycatch taxa by region. Source: NRC data files.
Bycatch regulation in the Northeast Pacific region began early in the century and largely involved efforts to minimize bycatch mortalities imposed by non-line fisheries on halibut. Following World War II, a rapid expansion of distant water fisheries occurred in the Bering Sea and Gulf of Alaska. These new fisheries were seen as a threat to U.S. and Canadian traditional salmon and halibut fisheries, and a number of bycatch agreements were established among the U.S., Canada, and other nations whose fleets fished off northwest North America. These agreements provided for the protection of (a) juvenile halibut grounds and (b) areas where U.S. fishermen traditionally fished crabs. In addition, the agreements led to an international observer program for distant water fishing vessels. U.S. groundfish fisheries expanded rapidly in the late 1980s, and a broad-based observer program for U.S. vessels was established early in 1990. The program has resulted in the development of a comprehensive dataset documenting discard levels for all groundfisheries of the region.
The Northeast Pacific discard data show current discard levels are relatively low for most species, although relatively high discard mortality is noted for halibut and a potentially significant problem may exist in the crab fisheries of the region. Impacts at the population level are relatively small for most species of groundfish, although altogether the total number discarded are estimated to exceed one billion annually.
Multi-species management options should examine selectivity and diversity measurements derived from discard and retained catch data. Values obtained through landings alone will mis-characterize the community effects of fishing in most instances.
