In scanning the literature concerned with bycatch and discards in world fisheries, it quickly becomes apparent the term “bycatch” has meant and continues to mean different things to different investigators. Furthermore, operational definitions of what is meant by bycatch are frequently not available in the published literature.
From our review of the literature, we conclude that “bycatch” has customarily been used to identify (1) species retained and sold, (2) species or sizes and sexes of species discarded as a result of economic, legal, or personal considerations, and (3) non-targeted species retained and sold, plus all discards. These terms may be further constrained in that within any one of these three categories authors may select and discuss only a portion of the bycatch caught with a particular gear type.
The first definition has been used by various investigators to differentiate target species from other species caught and retained. Australian authors have used the term “by-product” to refer to this form of bycatch. The second definition has generally been used by scientists reporting on fisheries of the Northeast and Western Pacific. A variety of scientists in other areas of the world have used the third definition which is consistent with that used by Saila (1983). In a few instances the term “joint catch” appears to have been used interchangeably with “bycatch” (Dewees and Ueber 1990).
Murawski (1992) states “the use of the term bycatch adds considerable confusion to a topic that is already complex to both scientists and managers”. He notes the term is relatively imprecise in that it constitutes a value judgment and may be inaccurate when used over any extended time to describe an element within a multi-species catch--in essence, “yesterday's bycatch may be today's target species.” During any specific fishing activity, the decision to set gear is often based on the expectation of harvesting a particular species. However, in other instances there may be no single target species in mind, the decision to fish being based on the anticipated value of the expected catch complex. Fishermen frequently make such choices based on the composition of past sets or the fishermen's collective experiences. It would appear the need to identify secondary target species as bycatch emerged from the desire of fishery statisticians to allocate fishing effort to different elements of a catch complex--perhaps justifiable from the researcher's requirements, but inappropriate in terms of the reality of many multispecies fishing practices.
Unfortunately, the term “bycatch” has taken on a much different connotation, in recent years. To many conservation, environmental, and fishery groups and to the lay public, bycatch is synonymous with “waste” and the catch of marine mammals and other “high profile” sea life.
Concerns with the terminology used to identify bycatch or discards were addressed at a bycatch workshop in Newport, Oregon (U.S.A) in February 1992 (McCaughran 1992). The authors have decided to use the following definition as proposed at the Newport Workshop as operational definitions within this monographs:
Target Catch -- The catch of a species or species assemblage which is primarily sought in a fishery, such as shrimp, flounders, cods
Incidental Catch -- Retained catch of non-targeted species
Discarded Catch -- That portion of the catch returned to the sea as a result of economic, legal, or personal considerations.
Bycatch -- Discarded catch plus incidental catch.
Incidental Catch Rate -- The Proportion of total catch which is incidental catch.
Discard Rate -- The proportion of total catch which is discarded. Rates may be computed for individual species or combined groups of species (Units of Measure--UM: kg/mt, numbers/mt, numbers/number, etc).
Discard Mortality Rate -- The proportion of the discarded catch that dies as a result of catching or handling processes.
Discard Mortality -- Discard mortality rate multiplied by discarded catch.
Prohibited species -- Any species which must, by law, be returned to the sea.
Unobserved Fishing Mortality -- Mortality imposed on a species by the encounter with fishing gear that does not result in capture.
High-Grading -- The discard of a marketable species in order to retain the same species at a larger size and price. The discard of a marketable species in order to retain another species of higher value. The retention of only those species or individuals within a species complex having the greatest market value; less valuable species or individuals are discarded.
Individuals Discard Quota (IDO) -- A quantity of prohibited species which is initially apportioned amongst individuals or groups of individuals to allow effective harvest of non-prohibited species. The IDQ may or may not be traded/sold within this group or between this group and other groups.
Discard Quota -- An amount of a prohibited species allowed to be caught by a particular gear type before constraining a fishery. It may be the sum of all IDQs if such a system were in place.
Discard Mortality Quota -- The discard quota multiplied by the discard mortality rate.
It should be noted that the definitions adopted for bycatch and discards are consistent with those used by Saila (1983). To the above list the authors have added the following operational definitions for terms used in this paper:
Discard ratio -- The ratio of discard to actual retained catch (e.g., discard/retained catch frequency given as a percentage).
Unobserved fishery mortality -- Death resulting from fishing that cannot be documented from observations of the on-board catch (e.g., deaths resulting from fish passing through webbing, freeing themselves from hooks, ghost fishing,etc.).
Black fish -- Non-reported catch.
Grey fish -- Catch which is misreported as to area or species
However, a litany of new terms is emerging in the bycatch literature which needs to be standardized. For example, terms identified in the Proceedings of the National Bycatch Workshop in Newport, Oregon, are similar to terms used in Europe, but some differences in nuance exist.
Discard Catch -- usually shortened to discards in European discussions, is widely perceived as a problem of fisheries management. For the most part, this is seen as an economic waste of the species which are a scarce commercial resource. The problem of discarding of endangered species had not been viewed particularly seriously until recently. However, this is rapidly changing with the “wall of death” stories acting as a catalyst. It seems helpful to partition any discussion into considerations of commercial discards and of mortalities generated on non-commercial and, in particular, on potentially endangered species. These latter categories are often referred to as non-target mortalities.
Additional terms not covered in Newport are in use in the Economic Community. For example, the question of what is the targets and what is the incidental catch is only resolved in Economic Community fisheries when the fishery operates under a legal derogation from the main regulations in force. Under the common fisheries policy of the EC, a number of derogation fisheries are allowed. These permit the use of smaller mesh sizes than the standard regulation, but often restrict the landed to certain minimum catch rates of named targets species or maximum landed bycatches of named bycatch species. A problem this creates of particular concern to commercial fisherman is the bycatch of small commercial fish in the industrial fisheries (fisheries catching small species such as sandeel (sandlance,Ammodytes spp.) for reduction to fish meal and oil). Thus, Industrial bycatch2 may deserve a category of its own.
Terms often encountered in the ETP tuna seine fishery include rejects(part of the catch that is not landed because of quality) and collateral mortality (similarly to unobserved fishing mortality). Clearly, considerable differences in the use of the term bycatch and its associated terminology prevail regional and a need exists, as the documentation and evaluation of bycatch and discards continues, to standardize terminology and observational modes.
Numerous entities have been contacted in a attempt to prepare an exhaustive set of references on discard rates and problems. These sources included individual scientists whose interest in bycatch was reflected in their publications in the subject area, national fishery entities throughout the world, international fisheries bodies (IWC, ICES, ICLARM, IATTC, IPHC, FFA, SPC, ICCAT, IFC, sources and others were contacted in order to secure qualitative and quantitative data on the character and magnitude of discards occurring in various fisheries of the world.
In, conducting our literature search, the multiple meanings of the term “bycatch” at times made it difficult to acquire information regarding discards, especially total quantities and the biodiversity of discards from particular fishing activities. After a period of data collection and literature search extending over more than two years, the authors have accumulated and reviewed an extensive literature base concerned with discards occurring in world fisheries and scanned thousands of abstracts and noted references concerning bycatch. The papers and articles gathered on bycatch have been placed in a special collection within the library of NRC on Seattle, Washington.
2 Industrial bycatch is the resulting from fisheries for species rendered into meal and oil.
Articles received by NRC were first reviewed for their relevance to the bycatch topics (discard levels) under consideration in this study. Relevant documents were assigned a unique reference number and logged into FileMaler®, a computerized library and information retrieval system. Reference numbers were stored in FileMaker®, along with eight separate fields for article and publication titles, author names(s), date and type of publications, keywords lists, species discussed in the article, and the location of the study area/fishery. Principle target and bycatch species discussed in the article were recorded in the species field. In the location field, FAO fishery region and specific sub-areas or countries within each region where the work was focused were listed. The design of the FileMaker® system permits searching in any one or more of the nine information fields. Search times for a specific reference in the database typically requires two or three seconds. Reference numbers assigned to each article FileMaker® were recorded on the top left-hand corner of the cover page of each hardcopy document when it was stored in the NRC library.
In our development of regional and global bycatch and discard estimates, the authors have reviewed some 820 papers pertaining to all aspects of the bycatch problem. To chronicle this review, a world bycatch database was created using a commercial computer spreadsheet program. Nineteen unique types of information were recorded in this database for each reviewed article. This information was recorded using discrete columns containing study year, FAO region, sub-region or area, fishing nation, gear type, target species or species complex, bycatch or discard species type, weight and/or number of discards, weight and/or number of the target species, and four calculated columns which derive discard per target species on a number per number, number per weight, weight per number, and weight per weight basis using data elsewhere in the subject row. Three additional columns denote whether the discard values refer to discards or more generically to bycatch, ratio of total retained weight or number to target species weight or number, and the reference number assigned to the article and recorded in FileMaker®. A final note field was used for short qualitative description of key characteristics of the target or discard data contained within the article.
The nature of computer spreadsheet programs and the structure of the world bycatch database permit a number of sorts, extractions, and other database manipulations allowing access to bycatch discard data of specific interest to a given user. For example, bycatch data pertinent to particular areas, gears, or species can be acquired by sorting by each of these fields individually. Likewise, discards of a given species by a certain gear type in selected areas can be reviewed by using all three of these fields when sorting. Total retained weight or number can be calculated by multiplying target weight or number by the ratio of total retained to target weight or number, Discard ratios for individual species within a fishery can be obtained directly from the database. Qualitative information recorded in the notes field can contain important insights into the quantitative discard data recorded for each fishery and should be consulted by each user (Figure 1 and Table 1).
The bycatch database emerging from this review currently contains more than 1,700 records and continues to grow as new documents are added to the NRC bycatch library. A sample page from the database is included as Figure 1. Table 1 contains a list of fishery, area, and species codes used in the database. Of the records in the database, 1,012 contain quantitative assessments of discards on an area/gear/target and non-target species basis. The remaining 711 records contain qualitative information pertaining to one or more of the many questions (economic, biological, ecological, social, legal, and ethical) surrounding the bycatch issue.
Bycatch and discard data reported for various regions and fisheries differ from fishery to fishery. Acquired target and non-target data are most frequently given in the form of numbers or weight and less frequently in terms of CPUE. Most authors report discards as a function of catch retained, e.g., the ratio of numbers discarded to numbers retained or the ratio of weight discarded to weight retained. Other report numbers retained per retained weight of target species. Studies classifying target and non-target catch by numbers and weight or providing a ready means of converting one measure into the other were rare. Recognizing the diverse characteristics of the recorded observation, the NRC bycatch database was designed to permit entry of either or both data types. A total of 727 weight-based records and 285 numbers-based records currently reside in the system. Of these records, 100 contain target and non-target data quantified in terms of both numbers and weight.
|Count||Period||FAO Region||Area||Fising Nation||Gear||Target Species||Discard Species||Discard Number||Discard Wgt (mt)||Retained Target Number||Retained Target Wgt (mt)||Discard Number per Target Number||Discard Number per Target Weight||Discard Weight (mt) per Target Weight (mt)||Retained Total Target Target||Source Reference Number|
Figure 1. Example records of the world bycatch database
|FAO Region||Area||Fishing Nation||Gear||Target||Bycatch Sp|
|1||See FAO||Bering Sea||US||Bottom Twl||Pollock-32||Juvenlle Chinook||1|
|2||Codes||Gulf of Alaska||Canada||Pelagic Twl||Pacific cod-32||Chinook||2|
|3||99=All||Brit. Columbia||Japan||Unsp. Twl||Groundfish||Coho||3|
|4||999=All US||West Coast||Taiwan||Gillnet||Squid-57||Sockeye||4|
|5||N.Pac High Seas||Korea||HS Driftnet||Other Salmon-23||Pink||5|
|7||Baltic||JV (US/For)||Purse Seine||Halibut-31||Unsp Salmon||7|
|8||Irish Sea||Iceland||Troll||Atka mackerel-33||Steelhead||8|
|9||North Sea||Poland||Sport||Rockfish-33||Atlantic cod||9|
|10||W. Greenland||Ireland||Seine (Danish)||Yellowfin sole-31||Herring||10|
|13||NAFO 3L||Germany||Beam Trawl||Sprat-35||Various Spp||13|
|15||NAFO 3M||Portugal||Dredge||Atlantic cod-32||Hake||15|
|20||France/Gulf of Lions||Greece||Redfish-33||Turbot||20|
|23||Celtic Sea||Holland/Netherlands||Lobster, Spiny-43||Greenland Halibt||23|
|24||Bay of Biscay||Australia||Lobster, Slipper-43||Redfish||24|
|25||ICES VI, VII, VIII||Faeroese||Lobster, American-43||Shark||25|
|29||Gulf of Mexico||EEC (ICES)||Herring-35||Lobster, Spiny||29|
|30||Atlantic Canada||Phillippines||Megrim||Lobster, Slipper||30|
|31||Southeast US||China||Mackerel-37||Lobster, American||31|
|33||E.Trop. Pacific||Sri Lanka||Sole-31||Marlin||33|
|35||Faeroe Islands||Arabian Gulf||Flounder-31||Turtles||35|
|38||Great Lakes||IATTC||Misc Flounder/Sole-31||Dogfish||38|
|40||ICES IV||Guyana||Atlantic salmon-23||Skate/Rays||40|
|41||ICES IVa||Mozambique||King crab-42||Sardines||41|
|43||Tasman Sea||SADDC (Southern Africa)||Whitefish-32||Mackerel||43|
|44||Sub-Trop Convergence||Senegal||Drum||Var. Bony Fishes||44|
|47||Sri Lanka||Peru||Castanha Family||Tanner crab/Other Crab||47|
|48||Indonesia||Virgin Islands||cateceens||King crab||48|
|50||Arabian Gulf||Sweden||Menhaden-35||Var. Crustaceans||50|
|51||Thailand||Crab, Unspecified-43||Pacific cod||51|
|53||Gulf of Paria||Greenland Turbot-31||Monkfish||53|
|55||New Caledonia||Shallow flatfish-31||Squid||55|
|56||West Central Africa||Sablefish||56|
|58||West Trop. Pacific||Weakfish||58|
|59||Gulf of California||Butterfish||59|
|61||Persian Gulf||Striped bass||61|
|62||Southern Oceans||Horse mackerel||62|
|63||South China Sea||Trout||63|
|65||Gulf of Mannar||65|
|68||Rio Grande del Sol||68|
|69||Gulf of St. Lawrence||69|
|72||West Central Atlantic||72|
Trawl fisheries account for 571 of the weight-based and 75 of our numbers-based entries. Net fisheries (driftnet, gillnet) are the second most common gear type in the database and account for 107 numbers-based and two weight-based records. Fifty-eight weight-based entries are linked to line fisheries. Line fisheries also contribute 33 numbers-based records. Weight- and numbers-based entries are more evenly split for pot gear, with 56 weight-based values and 41 numbers-based records. Table 2 displays the total number of database records and the number of records in weight- and numbers-based formats for each gear type. The “total” column involves records containing quantitative or qualitative information. For some gear types, the sum of the weight-based and numbers-based columns exceeds the totals column. This indicates that for this gear type a number of records contain both numbers- and weight-based ratios.
|Gear Type||Total Number Records||Number Records Weight-based||Number Records Numbers-based|
Each of the major FAO fishing areas is represented within the database. However, bycatch and discard data are best documented for fisheries of the Northeast Pacific and Northeast and Northwest Atlantic, and these areas account for 1,229 of the total database entries. In the Northeast Pacific, extensive observer programs initiated in 1990 have provided comprehensive catch and bycatch data for the groundfisheries of the region. Increasing levels of observer coverage in the crab fisheries of the Northeast Pacific are also contributing new information for this region.
In the Northeast Atlantic, ICES bycatch monitoring programs have provided extensive datasets covering the discards and industrial bycatch of the whiting (Merlangus merlangus), hake (Urophycis tenuis), haddock (Melanogrammus aeglefinus), and plaice (Pleuronectes platessa) and other demersal fisheries.
Moreover, the region has been the center of a great deal of bycatch reduction gear experimentation, and a number of papers dealing with technological solutions to discard problems are referenced in the database. Also, the Northwest Atlantic Fisheries Organization (NAFO) has published numerous studies dealing with bycatch in its subject region. In addition, the U.S. National Marine Fisheries Service (NMFS) and Canadian Department of Fisheries and Oceans (DFO) have added substantially to the documentation of bycatch off both coasts of North America.
Other important regional data contributions include FAO Areas 31 (principally the shrimp trawl fisheries of the Gulf of Mexico and U.S. Southeast), 51 and 57 (East and West Indian fisheries), 61 (Japanese and Chinese fisheries and the high seas driftnet fisheries of the region), 71 (fisheries of Southeast Asia, Micronesia, and northeastern Australia), 77 (purse seine tuna and pot lobster fisheries of the ETP), and 81 (line and trawl fisheries of the Southwest Pacific). Only spotty data records were identified for the Mediterranean (Area 37), the Eastern Central Atlantic (Area 34), the Southeast and Southwest Atlantic (Areas 47 and 41), and the Southeast Pacific (Area 87). A map of FAO fishery areas is shown in Figure 2.
Figure 2. FAO Fishery Statistical Areas. Source: FAO
A total of 55 target species/species groups and 63 bycatch species/species groups is represented within the database. Table 3 documents the number of records associated with the ten most frequently referenced target and bycatch species. Approximately 29% of the database records are associated with target fisheries for shrimp and groundfish species groups. Other groundfish species, such as Alaska pollock (Theragra chalcogramma), Pacific cod (Gadus macrocephalus), Atlantic cod (Gadus morhua), and whiting, contribute an additional 299 records. Fisheries for squid, Nephrops, tuna, and salmon round out the top ten target groups. Together, these ten target species/species groups constitute 1,158 records, two-thirds of all database entries.
|Target Species||Number Records||Bycatch Species||Number Records|
|Pacific cod||94||King crab||48|
|Atlantic cod||70||Tanner crab||41|
In contrast to the target species breakout, the top ten most referenced bycatch species account for just 47% of all database records. This result should be expected, given that a number of discarded species are often taken in association with a dominant target species. It is also clear bycatch research associated with species having high economic or social value is particularly prevalent in the literature. For example, records pertaining to salmon, halibut, and marine mammals together total almost twice the number of records for all other bycatch species in the top ten list and 30% of all records in the database.
3 If group species are disaggregated, rankings will change significantly.
The bycatch database offers a means of deriving estimates of a number of parameters regarding areas, species, and populations impacted by bycatch and discards. Although the authors have taken considerable effort to ensure that the global review is exhaustive, it is recognized important literature sources may not have come to our attention and there is a paucity of observations for some regions. Even for those areas where a great number of reported rates are available, the source data may be highly variable. Mangel (1993) discusses in some detail the preference for use of discard data in various forms, including CPUE numbers - and weight-rated based data, and the potential bias of datasets because of time/area variability or the application of certain statistical procedures to species having clustered distribution patterns. In this respect it is important to emphasize discard rates for a particular fishery usually involve a great deal of inter-annual variation, often related to recruitment patterns, changes in regulations, and fishery practices. Because the ratio-based information constituted the only database having adequate regional and global observations, it was chosen as the basis of our estimates. Even this data source does not provide adequate observations for calculation of statistical confidence levels. Thus, all estimates should be seen as provisional in character.
To address situations where discard data associated with a target species are lacking in a region, the authors have applied a set of criteria for selection of surrogate rate data. These bycatch rate selection criteria are:
Record must be post-1980
If a ratio exists for Target Species A in the region in question, use that ratio
If not, use a ratio for Target Species A from a neighboring region
Or apply a ratio for Target Species A derived from a global average of ratios associated with that species
Or apply a ratio for the species group within which Target Species A resides that is derived from a global average of ratios associated with that species group
Derive minimum ratios from summary records (e.g., with bycatch codes 13 and 99)
Use minimum-to-maximum range or, when applicable, the global average must include all gear types for which data are available
Estimates pertaining to the Bering Sea groundfish fishery are derived from comprehensive NMFS observer sampling data for 1992
In calculating global removals and in developing summary tables of discard ratios by gear or region, the authors have, in a few instances, used their best judgment in the selection of available bycatch data for particular regions or fisheries with which they are especially familiar, such as, the Northeast Pacific or the groundfish fisheries. In some cases, assumptions have been made about average weights of individuals discarded in order to convert numbers-based ratio data into a weight-based form which could contribute to global estimates of discards. These unit weights can be deduced by viewing formulas in database cells for discard weight corresponding with the subject target species/region combination. In order to improve upon all such estimates and keep the database current, the authors invite on-going discussion of these criteria, judgments, and assumptions from regional fishery experts from around the world. Suggested changes can be readily incorporated into the bycatch database.
Using the above criteria, the authors first analyzed the discard rate data (on a weight basis) considering target species and FAO regions, in order to arrive at an estimate of total global discards. Average annual commercial harvests by species and region for the period 1988 through 1990 were obtained from FAO's computerized landings database, FISHSTAT-PC.4 These volumes were then applied to the most recent maximum and minimum bycatch ratios recorded in the database for target species in each region. An estimate of discards associated with that target species was obtained by averaging observed maximum and minimum removal values for specific fisheries and regions. This process resulted in a mean estimate of 27.0 million mt of global discards5, based on a target catch of 77 million mt.6 Using minimum and maximum observations, we estimate a global discard range of 17.9 to 39.5 million mt.
Not included in these figures are data from freshwater fisheries and marine molluscan fisheries. Aquacultural harvests of shrimp and salmon have also been backed-out of the target harvest baseline. Recreational fisheries discards are also not included in these estimates. While reliable estimates of the weight of global recreational bycatch are lacking, data collected from U.S. fisheries in 1989 document a total discard of 1.035 billion individuals in association with a landed catch of 651.8 million individuals (1.5 individuals discarded to each fish retained). Clearly, recreational fishery discards would add substantially to estimates of global bycatch removals, and a definitive study of their extent is necessary. Further, subsistence fisheries harvest and discard rates were not available.
Our estimate of 27.0 million mt of discards is obviously much larger than Saila's (1983) value of 6.72 million mt. This higher estimate may be indicative of a worsening situation regarding discards, but may also reflect the growth of the available data regarding the scope of the problem. Saila (1983) focused much of his attention on shrimp fisheries. Mean total shrimp bycatch in the present study is estimated at 11.2 million mt, using more current data. 7 When adjusted for bycatch fractions which are retained, the discards decline to about 9.5 million mt, closer to Saila's estimate from 1983 (Table 4).
4FISHSTAT-PC is a computer program and database developed by FAO containing historical commercial catch data by region and species.
5 Includes bycatch landed but unreported by species in industrial fisheries.
6Some bycatch observations, particularly for tropical shrimp fisheries, did not provide estimates of the discard portion. Thus, the authors have arrived at discard estimates in these fisheries based on regional comments by numerous authors and inquiries to local fishery managers. Nevertheless, the estimates are at times subjective in nature.
|Area||Estimated Bycatch (mt)||Estimated Discard (mt)|
|West Central Atlantic||1,310,653||1,271,334|
|East Central Atlantic||123,636||61,818|
|Western Indian Ocean||1,871,075||748,430|
|Eastern Indian Ocean||482,879||289,727|
|West Central Pacific||1,450,352||1,377,835|
|East Central Pacific||590,955||561,416|
Throughout much of North and South America, a very large share of the bycatch taken during shrimp fishing is reported as being discarded. Similar situations are reported in the large Australian northern prawn fishery and in the shrimp fisheries of Europe. For Africa and some areas of Asia, a considerable portion of the bycatch is retained by artisanal fishermen. However, in many other regions the artisanal fisheries have been replaced or supplemented by modern shrimp trawlers and the percentage of bycatch retained has declined. Nevertheless, the trend may be offset by improved markets for species previously discarded in areas of Africa.
7Independent estimates of maximum total bycatch in the shrimp fisheries converge at between 16 and 17 million metric tons. Andrew and Pepperell's 1992 estimate of 16.7 million metric tons is only slightly higher than NRC's estimate of maximum shrimp bycatch of 16.4 million metric tons derived in this study.