Discards represent a significant proportion of global marine catches and are generally considered to constitute waste, or suboptimal use of fishery resources. A number of United Nations resolutions have drawn attention to the need to monitor and reduce discards and unwanted bycatch, in order to assess the impact of discards on marine resources and promote technologies and other means of reducing them. The previous FAO estimate of discards at a global level (referred to hereafter as the Alverson assessment), based on data prior to 1994, is considered to be outdated.
The present study re-estimated discards at a global level using information from a broad range of fisheries in all continents.
Selected policy and technical issues are highlighted and suggestions made for future actions. A road map for achieving further precision in the global estimate is described and associated initiatives are outlined.
The Alverson assessment is based on the use of the FAO Fishstat database of national catches. This database provides catch (in practice, the live-weight equivalent of landings) information by country, FAO area and species (or species group). The Alverson assessment is essentially a function of landings by species. However, there is no a priori reason why the discarded quantities of a species should bear a relationship to the landings of target species.
The approach used in this study is based on the premise that discards are a function of the landings of a fishery, rather than a function of the landings of a particular species. A fishery is defined in terms of an area, a fishing gear and a target species.
A list or inventory of the worlds fisheries was compiled in a discard database. Each database record contains quantitative data on: (i) the total landings of the fishery; and (ii) either the total quantity of the discards or the percentage of the total catch that is discarded. The total quantity of discards for a given fishery was generally extrapolated from the results of studies on a sample of the fishing activities.
The sources of the information on landings and discards are provided with respect to each fishery, so that the estimate can be readily verified, updated or changed, as new or more accurate information becomes available at national, regional or FAO level.
Discards (or discarded catch) were defined (FAO, 1996b) as being that portion of the catch which is returned to the sea for whatever reason. Post-harvest waste and discards of recreational fisheries are not included. Information on discards of turtles, seabirds and marine mammals is included in the database, but such incidental catches are a secondary target of the study. The study does not quantify either the unseen mortalities caused by fishing or the survival of discards.
The information contained in the database was compiled from three principal sources: (i) from scientific literature and from published national fisheries information; (ii) from reports and grey literature available within FAO or publicly available on the Internet; and (iii) from contacts with experts in national fisheries administrations, research institutions or regional fisheries organizations, many of whom provided detailed reports and databases.
The database contains four groups of fields:
those specifying the fishing area, which include reference to the FAO fishing area codes and the country or Exclusive Economic Zone (EEZ) from which the catch is reported;
those describing or designing the fishery, which include reference to the gear and the target species;
those quantifying the landings by the fishery and the quantity or percentage of discards - the sources of the quantitative information are cited;
other descriptive fields, which give the reasons for discards, relevant management measures in force, exploitation status of the fishery and other information of relevance to the analysis.
The fishery-by-fishery approach encountered several difficulties in data compilation:
the sheer scale of the task of compiling a list of the worlds fisheries and quantifying the landings of each one;
the absence or inaccessibility of information on discards for many fisheries;
the lack of published national fisheries catch statistics on a fishery-by-fishery basis;
the failure of numerous publications to distinguish clearly between discards and bycatch; and
the narrow focus of some studies on the discards of target or commercial species only.
To facilitate the discard estimates, certain assumptions were made, and use was made of fisheries information that had already been aggregated, specifically:
in the absence of information to the contrary, artisanal fisheries were assumed to have a discard rate of 1 percent or less than 1 percent of the catch;
in the absence of information to the contrary, fishmeal fisheries were assumed to have a discard rate of 1 percent or less than 1 percent of the catch;
with some exceptions, Southeast Asian fisheries were considered to have a discard rate of 1 percent of the catch;
tuna and other highly migratory species (HMS), and other fisheries for which statistical information has been collected by regional fisheries bodies (RFBs) were generally aggregated by ocean; and
fisheries, in the opinion of the author considered to be substantially similar in terms of fishing grounds, target species, fishing area, socio-economic basis and management regime, were considered to have a similar discard rate.
Over 2 000 records of fisheries were compiled of which 1 275 contain quantitative information on either landings or discards. Of these records, 788 are quantitatively complete, i.e. they contain quantitative information on both landings and discards for a given fishery. Countries with such complete sets of information include Norway, Iceland, the South Pacific Island states, Thailand, Malaysia and Viet Nam. In the case of the Southeast Asian countries this completeness is based on assumptions made by national fisheries authorities regarding low discard rates, rather than on empirical information on discard quantities. There are 62 records that refer exclusively to numbers of marine animals caught incidentally (marine mammals, seabirds and turtles).
Based on the set of complete records, the sum of the recorded discards is 6.8 million tonnes with respect to total recorded landings of 78.4 million tonnes. The global weighted discard rate is 8 percent.
Applying the global weighted discard rate estimated in this study (8 percent) to a ten-year average of the FAO Fishstat reported global nominal catch, total extrapolated discards are 7.3 million tonnes. Some caution is required in extrapolating from the total global catch, as certain major fish producer countries are not adequately represented in the database. These include the Democratic Republic of Korea, the Republic of Korea (no discard information), the Russian Federation, New Zealand and the Philippines. The European Union (EU) member countries and India have only partially been covered. A number of small fish-producing countries are not included.
Shrimp and demersal finfish trawl fisheries account for over 50 percent of total estimated discards while representing approximately 22 percent of total recorded landings. Tropical shrimp trawl fisheries have the highest discard rate and alone account for over 27 percent of total estimated discards. Small-scale fisheries generally have lower discard rates than industrial fisheries. Purse-seine, handline, jig, trap and pot fisheries have low discard rates. In geographical terms the highest discards are in the Northeast Atlantic and Northwest Pacific, which jointly account for 40 percent of discards (FAO areas 27 and 61, respectively).
At the global level it was not possible to compile a time series on discards to enable an empirical assessment of global trends in discards to be established. Nevertheless, two trends are apparent. There has been a reduction in bycatch and in discards in many fisheries, particularly those in developed countries. There is increasing utilization of bycatch and a consequent reduction in discards in many fisheries, particularly in developing countries. Several time series of discard data for selected fisheries are provided in support of these conclusions. A decrease in effort and change of target species in some major trawl fisheries has also resulted in a reduction of discards. Changes in fisheries regulatory regimes, requiring more selective fishing and prohibiting or curtailing discards, have also contributed to discard reduction.
The Alverson assessment, published in 1994, estimated discards to be 27 million tonnes (range 17.9 and 39.5 million tonnes). A subsequent (1998) FAO estimate suggested a reduced estimate of 20 million tonnes and a further study by Alverson in 1998 indicated that the 1994 assessment was an overestimate. Because of differences in method, the estimates provided in this report are not directly comparable with the Alverson assessment and consequently the extent to which the estimates represent a reduction in discards is not known.
The main spreadsheet file of the discard database and a bibliography are provided on the accompanying CD-ROM. The spreadsheet file is supplemented by numerous country and fishery files as well as files generated from databases supplied by the regional fisheries organizations or derived from national fisheries statistics. These files and source materials, including electronic copies of reference materials, are archived within FAO, classified by continent, country or regional fisheries organization. A searchable bibliography was compiled using bibliographic software.
The discard problem embraces several issues or subproblems:
the moral problem of responsible stewardship of marine resources;
designing a management regime that limits or prevents discarding while meeting multiple social, economic and biological objectives;
the practical problem of enforcing regulations designed to prevent or minimize discards, particularly as discards occur at sea where enforcement is most difficult;
the technical problems of gear selectivity and utilization of species with a low market demand through transformation or adding value; and
the economic problems posed by efforts to reduce bycatch, increase landing of bycatch or increase utilization of bycatch.
International instruments, including United Nations (UN) resolutions, the Kyoto Declaration and the Code of Conduct for Responsible Fisheries (CCRF) have highlighted the need to reduce or minimize discards. These instruments reflect the idea, enshrined in many of the worlds religious and secular beliefs, that wastage of natural resources is morally wrong.
A number of countries have instituted fisheries policies and management regimes based on the principle of no discards. A no-discard policy implies a paradigm shift in approaches to fisheries management. It moves the focus of management measures from landings to catches and from fish production to fish mortality. In conformity with the precautionary approach, by regarding no discards as the norm, any discarding then requires adequate justification.
Issues related to the Code of Conduct for Responsible Fisheries
There are two principal approaches to addressing the discard problem:
increasing utilization of bycatch
These two harvest strategies may be complementary and in any given fishery an appropriate balance between bycatch reduction and utilization is required. The biological and social principles upon which such a balance can be based require further analysis and development of decision frameworks. A more precise interpretation of the ecosystem approach in terms of the trade-off between promoting bycatch reduction and promoting bycatch utilization may be of value. In particular, the balance between highly selective fishing that targets one trophic level (or species) only, and less selective fishing that is likely to impact upon several trophic levels (or species groups), may require further attention to enable best scientific advice to be made available.
A third approach is to improve the survival of discards and animals returned to the sea. This is of particular importance with regard to species groups such as marine mammals, turtles, seabirds, lobsters and crabs.
Responsible fishing operations (in relation to discards and bycatch) can be based on the following principles:
making efforts to avoid unwanted catches - in particular, catches of endangered species and unwanted catches and discards that may reduce biodiversity or disrupt ecosystem function or integrity;
where catches of unwanted species, sizes or sexes are unavoidable, making efforts to find sui uses for such animals, and/or if there is a reasonable probability of survival, making efforts to return the unwanted catch to the sea;
taking measures to increase the survival of unwanted catch destined to be returned to the sea;
keeping records of discards, if required for management purposes.
The incidental catch and subsequent discard of charismatic, protected or endangered species, such as turtles, marine mammals and seabirds, are likely to have an increasing impact on fishing activities and trade in fish products. The absence of a neutral and internationally accredited mechanism for compilation of information on the incidental catches of many of these species and for examination and promotion of best practices in mitigation measures may impede rational discussion and development of solutions.
Discard information has a high inherent level of variability requiring high levels of discard sampling to give accurate assessments. On-board observer reports are considered indispensable for accurate estimation of discards. Relationships between discard rates and other variables (e.g. landings, duration of trip, length of trawl tow, market prices) tend to be weak. Consequently, raising or extrapolating discard estimates derived from samples to the level of the fleet or fishery may have a high degree of error. Accuracy is dependent on the design of an appropriate sampling protocol.
Discards account for a significant mortality in fisheries. For numerous reasons discard estimates may not be included in stock assessments, TAC determination or quota management. In general, the accounting toolkit for discards is deficient.
National fisheries statistics are generally collected, compiled and presented on a species-by-species or species group basis. There are several advantages in also compiling national fisheries statistics on a fishery-by-fishery basis. In particular, this may focus attention on the definition of coherent management units, link trends in landings to fishery-specific management measures and facilitate inclusion of discard estimates if required.
The discard database includes information on fishery management measures associated with discards and bycatch. The measures include legal obligations (e.g. minimum landing sizes, quotas and transhipment prohibitions), economic incentives and technical improvements (e.g. bycatch reduction devices [BRDs]). A number of fisheries have specific bycatch plans or require environmental impact assessments that specifically address bycatch and discard issues.
The development of guidelines on best practices can be considered with regard to:
discard sampling, e.g. from observers, logbooks, fishers estimates;
raising of discard estimates to the fleet or fishery level;
use of discard estimates in stock assessments;
use of discard estimates in total allowable catches (TACs) and quotas;
development of bycatch management plans; and
introduction and adoption of bycatch reduction and incidental catch mitigation technologies and practices.
A series of related studies can be considered to supplement this study, in particular, to compile:
information on the interaction between fishing activities and charismatic species at fishery, ocean and global level, with a focus on effective mitigation measures;
information on unobserved mortalities caused by fishing activities; and
additional information on survival of discards.
This study is regarded as an evolving tool rather than a static report. Ideally, it requires a further decentralized phase at national or regional level to: (i) verify or update the information in the discard database; (ii) give a broader ownership base to the discard information, through dialogue and consultation with national fisheries administrations and regional fisheries organizations; and (iii) compile discard information from countries and fisheries where information is deficient.
The global fishery-by-fishery records of landings form the backbone of the discard database. This set of records is of potential use for a range of other analyses, in particular if fields such as status of exploitation of the fishery are complete. Efforts are under way to integrate the database into FAOs Fisheries Global Information System (FIGIS) both as a basis for compiling the global inventory of fisheries and as a discard database subset. Records in the database may be biased towards discards, since many of these records are derived from discards literature.
 Alverson et al.,
1994. This publication is referred to hereafter as the Alverson
 Fishstat Plus (version 2.3) of 24 July 2003. The nominal catch value excludes marine animals and plants.