Tunas and billfishes are broadly distributed throughout the world's oceans, and in most cases fisheries are directed at stocks that range over numerous Exclusive Economic Zones (EEZs) and international waters. The international exploitation of tuna stocks by different fishing gears often makes it difficult to monitor the catches and to develop management protocols. In addition, the broad distributions, highly-migratory behaviour, and complex life histories of the tunas and tuna-like fishes make it very difficult to study them. As a result, despite the large catches, our understanding of tunas and tuna-like fishes is far from complete. Certainly, much less is known about the biology, physiology, and ecology of tunas and tuna-like fishes than of most other species of fish of comparable commercial importance.
Although the mechanisms are poorly understood at this stage, there is growing acknowledgement of the fact that abiotic and biotic environmental changes significantly affect the distributions, and perhaps also the productivity, of tuna stocks. It seems highly probable that observed changes in tuna populations reflect larger-scale changes in the pelagic ecosystem. Thus, it is important to determine the nature and extent of the impact of climate variability on both the pelagic ecosystems and the tuna stocks.
The tuna fisheries affect the target species, of course. In addition, they affect the species that comprise the by-catches and have more complex impacts on the ecosystem as a whole. The Precautionary Approach specifies the need to conserve the by-catch species and the ecosystem. On a species-by-species basis, the assessment of the impacts and the path toward reduction of the impacts is relatively straightforward. However, the approaches to ecosystem study, and then ecosystem management, are inherently difficult due to the complicated nature of the ecosystem per se, and also to the lack of consensus on the objectives of ecosystem management.
The biological peculiarities of the tunas and tuna-like species may severely affect stock assessment estimates, and consequently the validity of the traditional reference points, such as population and spawning stock size, recruitment, natural mortality, exploitation rates, etc. Moreover, these may have cascading negative effects that are difficult to handle, as often a minor part of these uncertainties is included in the models. Some of these are outlined here; however the priorities will be specific and driven by the stock assessment requirement (see Section 3).
In general, it is impractical to age adequate samples of tunas or tuna-like fishes directly. For this reason, age-structured assessments are typically based on conversions of length frequencies to age data, which presents various problems. Such conversions are difficult for tropical tunas that are more than about two or three years old because of their extended spawning seasons. Temperate tunas, which have more restricted spawning periods, can be aged to about four or five years with length-frequency data, but their life spans are much longer than that.
The feasibility and efficiency of direct ageing of tunas and tuna like species should be re-evaluated for both temperate and tropical tunas. Particular emphasis should be placed on improved size sampling by sex, improved statistical analysis of size distributions, validation of ageing methods, inter-laboratory calibration, and development of new methods.
Natural mortality is probably one of the major uncertainties, as all age-specific models require this information. Because of the present lack of data, many models consider a constant value of natural mortality for fish of all exploited ages, even if it is accepted by most scientists that it should be age-dependent.
Better estimates of age-specific mortality should be obtained.
Most of the biological parameters may vary substantially with time, in accordance with environmental variation and changes in exploitation rates. Some of them, such as growth and age, are density dependent. These should be more systematically studied because of their strong effects on stock assessment.
Estimates of key biological parameters should be revised periodically to monitor their temporal variation.
Tuna stocks are often distributed across entire ocean basins, and the same stock may be exploited by several different fisheries. The schooling and migratory behaviours of tunas make it difficult to sample them in accordance with rigorous statistical protocols. Most of the available information on relative abundance comes from catch rates for the various fisheries. Movement among areas is a key element of uncertainty in the understanding of the dynamics of tropical tunas.
Research on stock structure should be intensified, and should include development of improved genetic techniques and use of otolith chemistry techniques that have been used successfully for demersal fishes.
Priority should be given to large-scale tagging programs, using various types of tags, to measure the rates of mixing within and among management areas.
Spatially-structured models should be developed and tested.
Most stock assessments include the implicit assumption that an overfished resource will revert to its original status, the "virgin stock," if fishing is discontinued. It now appears, however, that severe overfishing can produce irreversible consequences, which may be due to elimination of one or more sub-populations.
The potential changes in genetic diversity of the tunas and tuna-like species due to exploitation should be studied.
As was noted above, tunas and tuna-like species are highly specialized, with a range of adaptations that allow them to respond rapidly to changes in their environment. Unlike demersal or coastal pelagic species, for which the ability to make large-scale migrations in response to environmental changes is limited, tunas and tuna-like species have the capacity to range over broad geographic areas in search of suitable habitat. Their highly-specialized visceral retia allow them to digest food very quickly, thereby ensuring that the maximum energy can be gained in suitable habitats.
It is likely that the productivity and distribution of a tuna stock is determined by the biological productivity of the pelagic ecosystem in which it is living. Climate variability affects productivity through changes of the physical processes in the ocean. Also, special features, such as hydrographic instabilities (currents or eddies) or biological enhancement due to run-off or atmospheric transport of elements, such as iron, which can limit the growth of phytoplankton, can be affected by the climate variability near large islands or continental masses. Therefore, variations in climate on various spatial and time scales can be expected to have large impacts on the spatial distribution and levels of abundance of tunas and tuna-like species. Identifying the mechanisms that control this environment-related variability is necessary to conduct the best possible evaluation of tuna stocks, to predict their short- and long-term variation, and to rationally manage the resources. Some examples of impacts of climate variability on tuna populations and fisheries are given below for different time scales.
The strongest and best-documented natural inter-annual climate fluctuation is the El Niño-Southern Oscillation (ENSO), which is an irregular low-frequency oscillation between a warm (El Niño) and cold (La Niña) states.
During El Niño events, the warm waters of the western Pacific (warm pool) extends far to the east in the central Pacific. Conversely, during La Niña events the warm pool is confined to the extreme western part of the equatorial Pacific. It has been demonstrated that such changes occurring during ENSO events have large repercussions on the movement and distribution of tunas. In particular, one of the most successful fishing grounds is located in the vicinity of the convergence zone at the eastern edge of the warm pool. This zone of convergence oscillates zonally over several thousands of kilometers in correlation with ENSO. The extension of the warm waters toward the temperate latitudes during El Niño events also appears to affect the migration of the North Pacific albacore, which are more dispersed in this extended habitat. Consequently, the catch rates are lower during such periods.
As a consequence of the eastward displacement of the warm water masses during El Niño events and the decrease in intensity of the upwelling, the thermocline becomes deeper in the central and eastern Pacific and shallower in the western Pacific. Adult yellowfin and bigeye apparently spend much of their time in or near the thermocline, with frequent ascending movements into the upper warmer waters. Given this behaviour, a deepening thermocline will increase the habitat layer, while a shallower thermocline will reduce it. Therefore, it is expected that a deeper thermocline (e.g. during an ENSO event) will decrease the success of purse-seine fishing in the eastern Pacific. Conversely, increases in longline catch rates observed in the eastern Pacific during ENSO events are generally explained by the deepening of the thermocline that would concentrate the fish in the deep layer where longlines are most efficient. However, it is difficult to discriminate from fishing data the changes due to the catchability from those resulting from spatial movements or variability in recruitment.
There is good evidence that ENSO events have an impact on the level of recruitment of the tunas in the Pacific. High (or low) recruitment episodes of South Pacific albacore and yellowfin have been proposed as the consequence of El Niño (or La Niña) events during the spawning seasons. For skipjack also, the fluctuations in the total catches by the purse-seine fleets in the western and central Pacific Ocean show an important interannual variability unrelated to the fishing effort. In particular, it is interesting to note that in the Pacific Ocean the mature phases of the two most powerful El Niño events, those of January 1992 and January 1998, correspond with the two lowest levels of catch for skipjack. Conversely, the situation was reversed during the 1996 and 1998-1999 La Niña events. These preliminary results suggest some opposite effects between the eastern and western Pacific that could be linked to the same mechanisms (wind stress, primary productivity, and temperature).
The effects of interdecadal climate forcing on the oceanic ecosystems have been clearly illustrated. One example is the (north) Pacific Decadal Oscillation (PDO). This oscillation is under the influence of the Aleutian Low-Pressure System, the intensification of which leads to stronger westerly winds in the central North Pacific. This climate change results in increased depth of the mixed layer and increased injection of nutrients into the euphotic zone from nutrient-rich deeper waters. The results suggest strong linkages between the atmosphere, oceanic mixing, and productivity across all the trophic levels of the marine ecosystem. Substantial impacts, with a range of variation in abundance of 30 to 50 percent, have been identified through most of the major components of the marine ecosystems-the primary and secondary producers, the forage species, and several levels of predators. Among the predators, it is likely that the North Pacific albacore is affected by the PDO, as suggested by the declining catches in phase with the environmental change.
Impacts due to long-term variability are obviously difficult to investigate because of the lack of good fishery statistics over long periods of time. Tuna fisheries have existed for several centuries in the Mediterranean Sea and near Japan. Historical data for the Mediterranean trap fishery show long-term fluctuations in the catches of Atlantic bluefin, by a factor 3 to 10, probably related to environmental variability, during the last two centuries.
The study of climate variability at these different time-scales is fundamental for predicting the potential impacts of the global warming on tuna stocks and, more generally, on the pelagic ecosystem. Simulations and scenarios of climate change due to greenhouse warming include increasing temperature, change in the illumination of the surface layer, where photosynthesis takes place, increasing stratification of the upper ocean, and changes in the oceanic circulation, reducing the nutrient input in the euphotic layer. All of these changes would have direct or indirect effects on the pelagic ecosystem, and thus on the stocks of tunas and tuna-like fishes. It is interesting to note that several simulations suggest that the changes in the mean state of the tropical Pacific Ocean would result in climate conditions similar to present-day El Niño conditions. The potential impacts on Pacific fisheries for tunas and tuna-like fishes include extension of present fisheries to higher latitudes, a decrease in productivity, mainly in the eastern Pacific, increasing variability in the catches, changes in the catchability of the different species, and increasing fishing pressure, particularly on bigeye and yellowfin.
From the previous review of studies on climate variability, it becomes apparent that the main mechanisms that link the climate variability to the biological changes in the oceanic ecosystem are similar, whatever the time scale. Scientists have synthethized these mechanisms in a conceptual model. Briefly, the primary forcing resulting from climate change is a change in surface wind stress that affects the depth of the surface layer and the horizontal and vertical flow and mixing within it. This leads to changes in primary production that are reflected in the secondary production. The growth and survival of members of the upper trophic levels are affected through effects on their larvae and juveniles, availability of food (affected both by changes in productivity and spatial distribution due to change of the meso-scale circulation), suitable habitat, and distribution and abundance of predators and competitors. After a regime shift, both top-down and bottom-up effects occur through the trophic ladder. Climatic variation acts essentially through its impact on success of recruitment. It also affects the fisheries through changes in the efficiency of the different gears. For example, purse seining is less efficient when the wind speeds exceed about 20 knots, and El Niño-related deepening of the thermocline in the eastern Pacific may reduce the efficiency of purse seines. It may lead to new fishing strategies, new technology, or transfer to new fishing grounds, e.g., displacement of the purse-seine fleets from Atlantic to the Indian Ocean and from the eastern to the western Pacific.
Research that leads to an understanding of the mechanisms underlying the recruitment of tunas and tuna-like fishes should be conducted.
Objective (empirical) methods for discriminating between changes in abundance, due to changes in recruitment, and changes in catchability and/or spatial distribution should be developed.
Definitions of by-catch, catch, discards, incidental species, landings, releases, and target species are given in Appendix 3.
From a global perspective, it is clear that despite the very large catches and significant value of fisheries for tunas and tuna-like fishes throughout the world, we have very little quantitative information on the nature and extent of the discards. This leaves open to question and speculation the impacts of these fisheries on populations of by-catch species. FAO has initiated projects on incidental longline catches of sharks and seabirds.
In the reviews following we summarize, for each major gear type, what is known about the by-catches, which include discards by tuna fisheries in various parts of the world. As few fisheries management bodies request that fishermen record records of discards in their logbooks, and when "sensitive" species are concerned it is highly unlikely that such data could be relied on, the data available derive principally from observer programs run by the CCSBT, IATTC, ICCAT, SPC, and national fisheries agencies.
By-and-large, the coverage of these programs has been limited to a small proportion of the fishing operations, over relatively short periods. Thus, the data collected provide an indication of the species composition in one or more of the capture components, but in most fisheries these data are insufficient to provide reliable estimates of the amounts of fish and other animals discarded dead or released alive, or trends in these.
The application of the Precautionary Approach in the absence of these data, particularly in the case where by-catch species have been classified as threatened, can lead to closures of fisheries. To avoid such drastic actions, we believe it is essential to improve the quality and quantity of observer and logbook data being collected, and to provide some objective assessments of the vulnerabilities of the by-catch species to over-exploitation. The level of observer coverage would have to be determined in accordance with the questions being addressed. For example, it has been calculated that coverage between 10 and 20 percent of trips is required to estimate the discard levels of the main by-catch species in the purse-seine fishery in the eastern Pacific Ocean.
The tendency, when looking at lists of by-catch species, is to ignore the species that occur infrequently, or in small numbers. For many species, such as albatrosses and turtles, however, their population sizes, life history characteristics, impacts on populations unrelated to the tuna fisheries, and other factors may all combine to indicate that even modest levels of mortality due to tuna fisheries could be of concern.
There are several characteristics that could be used to assess the relative vulnerability of a species, or group of species. These include, but are not restricted to:
characteristics associated with greater vulnerability
age at first maturity
reproductive life span
population status (difference between carrying capacity and population size)
characteristics associated with lesser vulnerability
litter size or batch fecundity
frequency of reproduction
difference between age of recruitment and first maturity
geographical range of the species in relation to that of the fishery
The challenge would be development of appropriate indices of vulnerability across phyla (e.g. birds, sharks, and teleosts). This could be achieved by applying appropriate weighting to the characteristics, however. Simulation modelling of any vulnerability estimation would be very useful.
While we remain largely ignorant about the impacts of tuna fisheries on by-catch species and pelagic ecosystems, it is obvious that these impacts have increased very significantly over the last 50 years as tuna fisheries worldwide have expanded their catches and effort by orders of magnitude.
Given our observations that data on the levels of by-catch species for most tuna fisheries will not allow quantitative assessment of trends in abundance of any species, let alone those that are vulnerable and/or present only in very small numbers, we are faced with the challenge of developing methods that will allow this assessment to be completed as quickly as possible.
This is a major challenge because, even if the levels of observer coverage were to be significantly increased, the distributions and catch levels of the by-catch species may not reflect their abundance. For example, in the case of purse-seine fisheries, the level and distribution of catch of species is likely to be influenced by the abundance, drift, and distribution of floating objects.
Research cruises could be planned to routinely estimate the abundances of some pelagic species, to develop some kind of baseline. If conducted on the ocean-basin scale of most tuna fisheries, a routine program that would adequately estimate the abundances of by-catch species for which the populations are small and diffuse would be extremely expensive. However, surveys at intervals of, say, 5 to 10 years, could provide valuable information on trends.
The international nature and basin-scale extent of most pelagic ecosystems remains one of the major obstacles to conducting this kind of research. However, the various RFBs and national organizations have given much lower priority to this type of research than to stock assessment. To effectively address such large-scale environmental problems, major changes in the policy and budget for each of these organizations would be required.
The longline fisheries, especially those of the high seas, catch large amounts of tunas, billfishes, other teleosts, especially pomfrets (family Bramidae), escolars (family Gempylidae), and moonfish (family Menidae), and pelagic sharks and rays, including blue (Prionace glauca), shortfin mako (Isurus oxyrinchus), oceanic whitetip (Carcharhinus longimanus), silky (C. falciformes), bigeye thresher (Alopias superciliosus), and pelagic thresher (A. pelagicus) sharks, all of which have some market value, and crocodile sharks (Pseudocarcharias kamoharai), velvet dogfish (family Squalidae), and pelagic stingrays (Pteroplatytrygon violacea) which have no market value. For various reasons, fishes other than tunas and billfishes may be retained and landed, or they may be released or discarded at sea. In some cases the fins may be removed from sharks and retained, while the rest of the shark is discarded.
Fish that would normally be retained must sometimes be discarded because they are badly damaged by sharks or mammals (e.g. Orca, Pseudorca).
In addition to fish, sea turtles, sea birds, especially albatrosses, and marine mammals are sometimes caught on longline gear. Sea turtles are hardy, and can usually be released alive. Albatrosses are caught mostly in temperate waters. Marine mammals are only rarely caught by longline gear.
The CCSBT requires that longliners use Tori pole streamers to reduce the catches of sea birds. In addition, experiments have been conducted with setting longlines only at night, use of devices to reduce the sinking time of the hooks, use of blue-stained bait, use of thawed bait etc. Use of artificial bait, variation in soaking time, and/or use of nylon monofilament leader may be effective for reducing the catches of sharks. Further experiments with hook timers or time-depth recorder may provide additional information that could be used to increase the selectivity of longlines.
Many small-scale drift gillnet fisheries for tunas and tuna-like fishes operate in the EEZs of many nations, especially in the Atlantic Ocean, Mediterranean Sea, and Indian Ocean. Little information on the by-catches of most of these fisheries is available.
Purse seines fish for skipjack, yellowfin, and bigeye in tropical areas and for bluefin in temperate waters.
Purse-seine sets could be classified in the following categories:
Schools associated with dolphins;
Schools associated with floating objects (either flotsam or FADs);
The species and sizes of fish caught by purse seines depend more on the types of sets than on the gear itself.
The dolphin fishery is directed at medium to large yellowfin, but incidental mortality of dolphins is a very sensitive issue in this fishery. This fishery provides a good example of how a by-catch problem has been mitigated by modification of the gear (the Medina panel) and techniques ("backing down," etc.). Sharks, and small amounts of various teleosts, are also caught with yellowfin associated with dolphins. Improved fish detection and identification techniques (i.e. long range sonar) could be useful for detecting tuna schools not associated with dolphins.
Sets on floating objects tend to catch smaller tunas and large amounts of by-catch species, including billfishes, sharks, dolphinfish (Coryphaena spp.), wahoo (Acanthocybium solandri), rainbow runners (Elagatis bipinnulatus), many of which are the targets of artisanal or recreational fisheries, and sea turtles.
Sets on unassociated schools catch tunas which are larger than those caught by sets on floating objects, but smaller than those caught by sets on dolphins. The species composition of the catches of by-catch species is similar to that of sets on floating objects, but lesser amounts of by-catch species are caught in sets on unassociated schools.
During the 1990s the proliferation of sets on FADs has changed substantially the fishing strategy of some purse-seine fleets, which has changed the species and size composition of the tuna catches. More than 50 percent of the tuna catches by purse seiners are presently taken on FADs. Large numbers of juvenile yellowfin and bigeye are taken under FADs, and, in some cases, discarded. The world-wide catch of by-catch species associated with FADs is estimated to be about 100,000 tonnes per year. The catch of by-catch species could possibly be reduced by modification of the fishing gear and/or techniques.
The extensive use of FADs could have other effects on the biology of tunas and tuna-like species. For example, the use of FADs might produce changes in the migration patterns, natural mortality, or growth of these species, or it might affect the associated fauna in ways that cannot be predicted.
In general, introduction of new gears and/or development of new fishing techniques that would increase fishing power should be closely monitored.
Experiments with sorting grids to facilitate the escapement of small fish should be conducted.
Attempts should be made to develop echo-sounders or sonars that indicate the species and sizes of the fish before they are caught.
The impact of catching small pelagic fish for bait by the pole-and-line vessels is probably negligible in coastal areas due to the large populations of those species, but it could be of some importance at offshore islands. The use of artificial bait would eliminate whatever objections there might be to catching small pelagic fishes for bait, and would allow pole-and-line vessels to fish more in offshore areas than they do now.
The possible impacts on bait populations in the areas where pole-and-line fleets are active should be studied.
Artificial baits for pole-and-line fishing should be developed.
There should be collaboration among tuna research and management organizations and others (other international organizations, non-governmental organizations (NGOs), and interested scientists) to estimate the bycatches, assess the stocks, and investigate ways to reduce the by-catches, especially when the negative effects are serious.
Managerial options to reduce the by-catches, such as requirements that all catches be retained, establishment of quotas on by-catches, establishment of upper limits for by-catch to catch ratios, areal and/or seasonal closures, and requirements regarding gear configuration and/or fishing techniques, should be investigated.
For many years fishery scientists and managers have relied on single-species models for managing fisheries. These models have been useful, but deficiencies are apparent.
Conceptually, the main problems are:
The models do not take into account the interactions among different species that are targets of different fisheries or are taken incidentally by them;
The models did not adequately account for impacts of the fisheries on the habitat, or on other species that occupy the habitat.
The problems with the catch data became apparent when observers first accompanied vessels during their fishing operations, and there was an opportunity, for the first time, to estimate the amounts of fish discarded and to incorporate these data into estimations of fishing mortality.
Some highly-visible discards of by-catch species, e.g. dolphins in the tuna purse-seine fishery of the eastern Pacific Ocean, sea turtles in the trawl fisheries for shrimp, sea birds in the longline fisheries for tunas and billfishes, and many species in drift gillnet fisheries, brought the environmental impacts of these fisheries to the attention of the public. Reduction of the discards was not the only thing that was needed, however. It was already obvious to many scientists that the effects of climate variability, the interactions between target and by-catch species, and the ecological impacts of the fishing operations should be incorporated into a unified system of fisheries management. Ecosystem management, or better, ecosystem-based fishery management, became a cliché repeated by all, but it was never fully defined as a desirable goal, and it is operationally unattainable today.
If, as the Precautionary Approach suggests, fisheries management should evolve in such a way as to require increased consideration of by-catch species, including ecologically-related species, and the ecosystem. It seems likely that this process will be gradual, and will be along the following lines:
Step 1: Modification of current single-species management to accommodate the Pre-cautionary Approach;
Step 2. Determination the appropriate ecosystem parameters, in conjunction with the development of ecosystem-based management goals, objectives, and schemes;
Step 3. Implementation of the Precautionary Approach to the ecosystem-based schemes.
In the context of this process we need to discuss the role of research, understanding that this is dependent on the managers' goals.
It seems likely that the general objective of ecosystem management will be to maintain ecosystem structure and function. Some specific goals are likely to be:
The impacts of the fishery on other species (discards, mortalities caused by the gear (retained or not retained by the gear), fate and significance of discarded biomass, etc.) are accounted for, and are maintained within pre-specified limits.
The overall capacity of the fishery is determined by the levels of available targets, and maintained within pre-specified limits.
The harvest strategies for individual species take into account the potential impacts of the fishery, and also the impacts of other significant anthropogenic factors, such as habitat degradation.
When many components of an ecosystem are being exploited, the harvest strategies for each exploited species are consistent with ecosystem goals.
Negative impacts on the habitat, such as discards, damage to the bottom and bottom communities, "ghost fishing," and pollution, are routinely assessed and, if needed, subjected to mitigation programs.
In general terms, ecosystem structure and function is maintained by the conservation of all stocks present within acceptable levels of biomass, and the maintenance of the existing dominance structure. Thus, research must be conducted to determine the properties of the ecosystems, such as the proportions by trophic level and guild, size-frequency distributions, and P:B and other ratios that should be monitored.
It must be acknowledged that we have yet to define or measure the critical parameters of pelagic ecosystems, and that we do not know how to do so.
In general, the catches of most of the world's tuna fisheries have been increasing over the last 10 years, and new techniques, such as FAD fishing, have been developed to increase the efficiencies of the vessels.
The world's tuna fisheries currently catch over 4 million tonnes of tunas and tuna-like fishes each year, and the exploitation rates for many of the stocks of these are estimated to be quite high. It is estimated that most tunas and tuna-like fishes have suffered significant decreases in their biomasses during the last 50 years.
Pelagic ecosystems inhabited by tunas and similar species, such as billfishes, can be characterized by their very large size, compared to any other ecosystems exploited by fisheries. The longline fisheries for tunas ands billfishes operate in an area of more than two thirds of the world oceans (about 50 million nautical miles2), while the surface fisheries for tunas are exploiting a smaller (but still very wide) area of about 20 million nautical miles2. These fisheries are exploiting the epipelagic layers of the ocean, the first 100 or 200 m for purse seiners and up to about 300 m, or even deeper, for the longline gear. Tunas, billfishes, and some sharks are among the few species that are heavily exploited in offshore pelagic waters. Many other pelagic species, such various other bony fishes, some sharks and rays, crustaceans, mammals, and molluscs) also inhabit these areas, but there are few fisheries directed at them.
The most significant recent change affecting ecosystems occupied by tunas and tuna-like fishes has been the increase in the use of FADs.
Multidisciplinary research should be conducted on FADs in the various oceans.
The principal goals of this research should be to obtain an understanding of the behaviour of the various species and sizes of fish and other animals associated with FADs and the functioning of this small-scale ecosystem to determine whether FADs are altering the movement patterns and biological characteristics, such as growth and natural mortality, of tunas and other species associated with the FADs. This research should be accomplished with at-sea surveys, experiments on FADs, and analyses of detailed data obtained from purse seiners and FAD supply vessels.
Impacts of the removal of tunas and tuna-like fishes
The effects of annual removal of over 4 million tonnes of tunas and tuna-like fishes from pelagic ecosystems are not understood. As these species are generally highly-productive high-level predators, we would expect, on the basis of trophic models, that the effect would be significant. However, empirical data on food-web effects are not available for pelagic systems, and most classical fishery models were not developed to address that subject.
Improved ecosystem models that will make it possible to better understand and predict the impact of removals of tunas and tuna-like fishes on pelagic ecosystems should be developed. In addition it would be useful to compare the structures of pelagic ecosystems in which the exploitation rates of tunas and tuna like species are know to be different.
Impacts of the removal of by-catch species
We do not currently understand the effects of the removal of by-catch species on tunas and tuna-like fishes. These species should be incorporated into the models discussed above. This will not be possible, however, until adequate data on the quantities of by-catches taken are collected. Also, we have little or no information on the relative abundances or biomasses of many components of the pelagic ecosystem.
Research should be carried out to:
Develop statistically-designed observer programs (or alternative schemes) to quantify the by-catches in each of the fisheries for tunas;
Identify the factors leading to by-catch mortality, and use the results of this research to devise mitigation programs;
Study, through a variety of approaches, such as modelling, simulation, and experimentation, the impacts of catches and by-catches on the ecosystem, with the objective of setting reasonable policies to use in the determination of the target levels for the different components of the ecosystem;
Develop a vulnerability index that will allow objective assessment of the potential impacts of the fisheries on by-catch species;
Develop monitoring systems to track changes in the communities affected by the fishery, with special reference to by-catch species, and to species receiving subsidies from the fishery (i.e. discards and fish hooked on longlines). The available options include estimation of abundances by periodic fishery-independent surveys or any other accepted methodology and the development of indices of relative abundances to monitor trends.
When there is scientific evidence that suggests that a species is vulnerable or endangered, it is important to understand the impacts of further removal of the species.
Impacts of the discarded biomass on the system
A fraction of the discards is consumed either on the surface or in the upper water layer by species such as dolphins, sharks, and seabirds. There is some indication that some species benefit from these subsidies, and that they expand their populations at the expense of others. In addition, several species of marine mammals, sharks, etc., have learned to take fish from fishing gear, occasionally getting entangled during the process. The fisheries can provide a competitive advantage to a species that breaks the balance reached through evolution.
The remaining fraction of the discarded biomass (of both target and by-catch species) sinks down the water column and ends up on the bottom in abyssal depths. The amount in question is not large, given the large area where it is dispersed. However, as the dumping may be quite localized, it may have local effects on abyssal communities.
Responses of tunas and tuna-like fishes to natural and anthropogenic changes
Understanding and ultimately predicting how populations of tunas and tuna-like species and the by-catch species respond to natural and anthropogenic changes is a major challenge for integrating the ecosystem approach into management. Modelling studies should be directed at elucidating the mechanisms linking biological and physical components of marine ecosystems and understanding the responses of the ecosystems, particularly the populations at higher trophic levels which are exploited by various fisheries, to various types of physical forcing and biological interactions. This would require the development of coupled physical-biological interaction models on the scale of ocean basins, and would include zooplankton, micronekton, and higher-level predators that are not exploited. This multi-disciplinary theme is currently being developed mostly by the international Global Ocean Ecosystem Dynamic (GLOBEC).
The linkages among physical conditions, marine food web structures, and ecosystem or population dynamics should be identified.
Networks of collaboration to take advantage of programs involved in ecosystem research should be developed.
Integration of climatic data into the models
Meso-scale hydrographic features in the ocean influence the distribution and survival of plankton and fish larvae. The development of climatically-driven physical oceanography models appears to be a modelling tool particularly well adapted to the dimension of the pelagic ecosystem for identifying advection routes and areas of retention of tuna larvae and juveniles. Such models should be useful for investigating the mechanisms of mortality at the different juvenile stages, so that the factors that lead to low mortality of larvae and juveniles, and thus maximize year-class strength, could be identified.
Biological productivity in the pelagic zone is highly dynamic, characterized by advection of organisms at lower trophic levels and by extensive movements of animals at higher trophic levels, both of which are strongly influenced by climatic variability. As the potential energy of stored biomass is transferred by the trophic level up the food web, the time scale for transfer between levels increases from seasonal to multi-annual scales. Ecosystem models coupling physical and biological interactions may be constructed from combinations of production and age- (size-) structured population models that are appropriate for simulating space and time distributions of successive cohorts of targeted species. A major research challenge of these modelling studies concerns the parameterization of the mechanisms of interaction among the different modelled components.
Single- or multi-species structured population models developed with rigorous statistical estimations relying on fishing and tagging data are useful for estimating the population parameters, such as biomass, mortality, growth, vulnerability to capture, and gear selectivity. Biological studies (growth, reproduction, and diet), physiological experiments, and acoustic tracking and archival tagging make it possible to better define various aspects of the biology of the species, including their habitat and behaviour. Individual-based models (IBMs) are a useful approach for integrating these different types of information and testing and formulating mathematically the behaviours, growth variability, and survival probabilities of individuals of specific species through time, as determined by environmental conditions. These formulations and parameterizations can be then included in structured population models, coupled to spatially-resolved ecosystem models for providing spatial distributions of the population studied which vary with time.
Transport models with appropriate biological components to determine how large- and meso-scale circulation patterns transport and influence the mortalities of larvae and juveniles should be improved and, if possible, automated.
The methods for sampling tuna larvae should be improved, and methods for sampling and identifying tuna eggs should be developed.
Although various area-time closures have been used for the management of tuna fisheries in recent years, the impacts of these are not yet fully understood. For example, short-term closures to fishing on floating objects have been implemented in both the Atlantic and eastern Pacific Oceans to reduce the fishing mortality of juvenile tunas.
In other areas of the world, however, permanent closures of areas have been established to protect the resources and the ecosystem. Even though tunas are highly migratory, there are areas that consistently show high concentrations of juveniles, aggregations of spawning individuals, or large by-catches of other species. As fishing effort could be redistributed outside the closed areas, and enforcement of the closures could be relatively easy with either observers or vessel monitoring systems (VMSs), this type of management measure should be considered as a potentially useful option to be added to those already available. It should be noted, however, that in many cases juvenile tunas are exploited in nearshore waters by artisanal fishermen, and it would not be easy to find employment for these fishermen in other fisheries.
Using models and empirical studies, it should be determined whether closure of areas affects the structure and integrity of pelagic ecosystems and the dynamics of target and by-catch species.
Studies that will form the basis for decisions on the selection of areas for closure, i.e. optimal sizes of the areas, periods during which they would be closed, and ecosystems that would be affected should be conducted.
The potential for use of remote-sensing data to predict areas of high concentrations of juvenile tunas and/or by-catches species should be determined.
Because many fisheries have been operating for decades, or even centuries, their impacts on the ecosystem may have become part of the evolution of those ecosystems. Equilibria prevailing prior to the fishery may not be reached again, even if the fishery, or its impact, is eliminated, as ecosystems do not necessarily go back along the same tracks they used to arrive at the current situation. This potential lack of reversibility in the dynamics of ecosystems may be due to various additive or independent causes, such loss of ecological niche, genetic erosion of individual populations, loss of specific biodiversity in the ecosystem, loss of keystone species, etc. The recovery of a population, or the reversal of ecosystem changes, depends on a series of factors, many of which are not under the control of man. This should not be considered as a justification for inaction, however.