Daniel Pauly* and Rainer Froese t*Fisheries Centre, University of British Columbia, and t International Center for Living

Aquatic Resources Management, Manila

fish spawning usually suffer tremendous and largelyunpredictable mortalities, thus uncoupling spawn-ing from recruitment.

trophic level A number indicating the position of aspecies within an ecosystem though the number ofsteps linking it to the plants. By definition, plantsare TL = 1, herbivores are TL = 2, and so on. Notethat trophic levels do not need to be whole numbers;intermediate values occur among omnivorous con-sumers.

I. Major Adaptations of FishesII. Respiratory Constraints to Growth and Related

ProcessesIII. Distribution of Exploited Fish StocksIV. Ecosystem Impacts of FisheriesV. Managing Fish Biodiversity Information

VI. Preserving Fish Biodiversity

GLOSSARY

FISH STOCKS ARE POPULATIONS OF "FISH," THATIS, VERTEBRATES WITH GILLS AND FINS, SUB-jECTED TO EXPLOITATION BY HUMANS. Popula-tions are components of species, inhabiting part of theiroverall range, and usually having little genetic exchangewith adjacent populations. The major adaptations de-termining the spatial distribution of fish stock biomasspertain to the anatomy, reproductive biology, and respi-ratory physiology of the species to which the stocksbelong. Also, fishing has become increasingly importantto the biodiversity of fish, either through its direct im-pacts (changes of stock size and age structure, andoverall biomass reductions, down to extirpation of pop-ulations), or by modifying the ecosystems in whichthey are embedded. Research devoted to monitoringthe biodiversity of fish (or other organisms) must beable to handle large amounts of suitably formatted dis-tributional information, here defined as consisting of

biomass Collective weight or mass of all the membersof a given population or stock at a given time, or,on the average, over a certain time period.

bioquads Occurrence record of organisms, serving askey units for biodiversity research and consisting offour elements (species names, location,. time, andsource).

catches The fish (or other aquatic organisms) ora givenstock killed during a certain period by the operationof fishing gear(s). This definition implies that fishnot landed, that is, discarded at sea, or killed by lostgear ("ghost fishing"), should be counted as pan ofthe catch of. a fishery.

ecosystem Area where a set of species interact in char-acteristic fashion, and generate among them biomassflows that are stronger than those linking that areato adjacent ones.

recruitment Entry of juvenile fish into the (adult)stock. Recruitment is distinguished from reproduc-tion, because the eggs and larvae that result from

801Encyclopedia of Biodiversity, Volume 2

Copyright @ 2001 by Academic Press. All rights of reproduction in any form reserved.

.

802 FISH STOCKS

"bioquads." Management regimes aiming at preservingfish biodiversity will have to include much stricter regu-lation of fishing and the establishment of no-take areas.

I. MAjOR ADAPTATIONS OF FISHES

A. Anatomy and Physiology

With about 25,000 recognized species in over 500 fami-lies, fish are the most diverse vertebrate group. How-ever, their watery habitat, while failing to protect themfrom modem fishing gear, makes it difficult to fullyappreciate this diversity, and the extent to which it isnow threatened. It is even more difficult for us, as air

breathers, to perceive the constraints under which fish,as water breathers, were forced to evolve.

Water is an extremely dense medium, 775 timesheavier and 55 times more viscous than air. Also, watercontains 30 times less oxygen than air, and this oxygendiffuses 300,000 times more slowly than in air. Thesephysical constraints, which shaped all early life-forms,including the jawless predecessors of the fish, the agna-thans, are best visualized by describing the major evolu-tionary trends leading from agnathans to modem fish(Fig. lA).

The first of these trends was the evolution of jawsfrom the first upper and lower gill arches of agnathans.This built on the intimate connection, in the most prim-itive vertebrates, between the feeding apparatus (i.e.,

Thunnus obesus, A = 7.5

Leiognathus dussumieri

t Pterolepis nitidus (15 cm)

Cyprinus carpio, A = 2.2

Gazza minuta

Pomatochistus minutus, A = 0.6

Secutor ruconius

Cyprinus carpio (1m) Cetorhinus maximus (15m)

FIGURE 1 Major evolutionary trends from agnathans to extant fishes. (Note that no direct ancestor-descendant relationshipsare implied among the groups depicted.) (A) Trends toward larger gills; (B) trends toward efficient jaws; (C) trends towardeffective paired and unpaired fins. [Note the aspect ratio of the caudal fin, defined by A = h2/s, where h is the height and s thesurface (in black) of the caudal fin, and of which high values define fast, large-gilled continuous swimmers, and conversely forlow values.)

803FISH STOCKS

FIGURE 2 Schematic representation of how changes in water levelcan multiply, by creating isolated subpopulations, the number of speciesoccurring in a given area. Such a mechanism, driven by repeated climaticchanges, is thought to explain the large number of fish species inSoutheast Asian marine waters and in the Great Lakes of Africa.

the mouth) and the respiratory apparatus (i.e., the gillsadjacent to slits on both sides of the anterior part ofthe alimentary canal). Water-breathing invertebrateslack this close connection between feeding and breath-ing, one reason why even the largest among them (giantsquids) cannot reach the mass of the largest fish (20metric tons, for the whale shark Rhincodon typus).

The reorganization of the head of early fish allowedlarger gills to evolve, which allowed the higher meta-bolic rates required for swimming in open waters. Thistransition was assisted by the gradual loss of the heavyarmor protecting the slow, bottom-slurping agnathans.The fine "teeth" covering the bodies of sharks are ves-tiges of this armor.

Fast swimming in open water required better fins,both for propulsion and for steering. Propulsion is pro-vided in most fish by oscillations of a caudal fin whoseaspect ratio (Fig. lC) gradually increased toward tunasand other derived, fast-swimming groups with verylarge gills. Steering, on the other hand, is provided bydorsal, pectoral, and anal fins. These fins are stiffenedfor precise action by hard, bony rays in the most derivedfish, the teleosts, whose evolutionary success was fur-ther enhanced by a complexly built protrusile mouththat enables capture of a wide range of food items(Fig. 1B).

Subtle anatomical changes in fish can thus createmore niches for increasing the numbers of specialists,which then occupy increasing numbers of closelypacked niches. Ecosystems in which these changes haverun for long periods, undisturbed by physical changes,therefore contain very large numbers of fish species.Their numbers are even larger in areas such as the GreatLakes of Africa and the tropical Indo-Pacific, wherechanges of water levels have repeatedly isolated basinsand subpopulations, thereby accelerating species differ-entiation (Fig. 2).

larvae encounter, even during spawning seasons at-tuned with zooplankton production cycles, are usuallyfar too low to allow survival of fish larvae, and theoverwhelming majority of such larvae perish. Thosethat tend to survive usually happened to have hatchedwithin plankton-rich water layers. These layers are usu-ally a few centimeters thick and last for only a few daysof calm, between wind-driven or other mixing events,such as storms or upwelling pulses, that enrich surfacewaters with nutrients from deeper waters. This impliesthat large biomasses of fish can build up only whenand where the local oceanographic conditions take theform of "triads" defined by (1) nutrient enrichment,such as generated by wind-driven mixing, (2) highplankton concentration, such as generated by variousmechanisms including fronts, and (3) retention of lar-vae, required to prevent these weak swimmers fromdrifting away from suitable habitat. In pelagic fishesthat build high biomass, for example, the anchoviesand sardines in coastal upwelling systems off northwest-em and southwestern Africa, Peru, and California, thesetriads occur only when the coastal winds range from 4 to6 m per second. Weaker winds do not generate enoughenrichment, and stronger winds disperse the larvae off-shore.

B. Reproduction and Recruitment

Though many ancient fishes such as sharks and raysor the coelacanth Latimeria chalumnae practice internalfertilization and produce few large eggs or live offspring,most recently evolved fishes produce numerous smalleggs that are fertilized externally and develop as partof the plankton, without parental care. The larvae thatemerge from those eggs, after less than one day in warmtropical waters and up to two weeks (and more for largereggs) in cold temperate waters, are usually elongated, asbefit small, finless zooplankton feeders.

The average zooplankton concentrations that these

804 FISH STOCKS

declines with size, because the two-dimensional gillarea cannot keep up with the three-dimensional in-crease of body mass. Hence larger fish dispose of rela-tively less oxygen to supply their metabolism, the rea-son why they ultimately stop growing. Also,environmental factors that tend to increase metabolicrate-especially elevated temperatures, but also includ-ing other form of stress-have the effect of reducingthe maximum size that the fish of a given populationcan reach (Figs. 3A and 3B). This is why tropical fishtend to be smaller than their respective cold-water rela-tives. A similar mechanism explains the nearly constantrelationship in fish between size at first maturity andmaximum size (Figs. 3C and 3D).

Fish have developed several strategies to deal withthe uncertain recruitment that results from the triadrequirements. One is being small, shott-lived, and capa-ble of quickly building up large biomass under favorableenvironmental conditions. The other is being large,long-lived, and capable of weathering long series ofrecruitment failures through repeated spawning by old,large, and highly fecund adults. An example of theformer strategy is provided by the Peruvian anchovyEngraulis ringens, whereas the northern cod, Gadus mor-hua, provides an example of the latter. Yet anotherstrategy is to reduce the dependence on environmentalconditions by various forms of parental care, such asnesting and guarding (e.g., in catfishes, family Clarii-dae), mouth-brooding (e.g., in cardinal fishes, familyApogonidae), and live-bearing (e.g., in ocean perches,genus Sebastes).

Another imponant feature of fish stocks is that, con-trary to earlier assumptions of homogeneity, most ap-pear to consist of well-differentiated individuals, eachaiming to reproduce at the very place where it washatched. Or, put differently: most migratory fish tendto "home." This behavior, well documented only inPacific and Atlantic salmon (Oncorhynchus and Salmo,respectively), implies that individual fish, when repro-ducing, do not seek "optimal" sites, but rather spawnas close as possible to the site at which they hatched,and to which they are imprinted. This tendency to eitherstay in or return to a certain area makes it difficult forfish stocks to rebuild once they have been decimatedby local overfishing or pollution.

B. Adaptation to Respiratory Constraints

Fish have ~volved various strategies and tactics to over-come respiratory constraints. One strategy, illustratedin Fig. IB, is to evolve large gills, a route taken bynumerous open-water ("pelagic") species, culminatingin tunas (Fig. 4).

Another strategy is the evolution of life cycles inwhich the juveniles migrate to deeper, cooler waters asthey grow and then, upon maturing, produce eggs thatquickly float up to the warmer surface layers, out ofreach of the often cannibalistic adults. Such typicalcycles are completed by an onshore drift of the larvaeto coastal areas, and productive shallow nurseries forthe early, voracious juveniles, which again migrate intodeeper waters as they grow.

A tactic to accommodate metabolic stress, which isparticularly useful in areas with strong seasonal temper-ature oscillations, is for the feeding adults to store fatduring the warmer part of the season (late summer toearly fall). Fat requires far less oxygen for maintenancethan protein of muscle and other tissues. As tempera-ture declines, the accumulated fat is converted intoother tissues, notably gonads, whose contents are shedin spring, thus reducing body mass when temperaturesagain start to increase. These cycles, which use fat asprotection against respiratory stress, are the reason whytemperate fish tend to contain more muscle and visceralfat than tropical species, where temperatures, althoughhigh, do not fluctuate much in the course of a year.

Another tactic that delays respiratory stress is associ-ated with ontogenetic shifts in diet composition. Here,the young fish feed on a diffuse, small prey (e.g., inverte-brate zooplankton), while the adults, via their sheersize, can capture energy-rich prey such as other fish,which are acquired at lesser cost by the predator.

II. RESPIRATORY CONSTRAINTS TOGROWTH AND RELATED PROCESSES

A. Basic Geometrical Constraints

Fish growth, as in other animals, requires both foodand oxygen, the latter being required to synthesize thesubstance (adenosine triphosphate or ATP) that servesas fuel to all organisms. For oxygen to be metabolicallyavailable, it must be inside the fish body, that is, it musthave passed though its gills. Thus, since oxyg~n cannotbe stored inside the fish body (contrary to food, whichcan be stored as gut contents and as fat), the metabolicand growth rate of fish are largely proportional to thesurface area of their gills. So fish that quickly reachlarge sizes have gills with large surface areas (as intunas), and conversely in slow-growing fishes (likegroupers). Moreover, gill area per unit of body mass

FISH STOCKS 805

A B

Wmax Wmax2 Wmaxl

c D

.:E0>

"CD"~>-

"U0.eI::0

"-E:.E:J0I::

8I::

~~0

'0.:E0>

"~>-

"U0

';ij~tU

aQ

Q

Wm1 Wmax1 Wm1 Wmax2

Body weight

Maintenance metabolism

FIGURE 3 Schematic representation of the relationships linking, in fish, respiratory area (and hence metabolicrate), maximum body size, and size at first marurity. (A) As body size increases, gill area per body weightdecreases, down to a level when it suffices only for maintenance metabolism. This defines the maximumsize that can be reached. (B) Any environmental factor increasing oxygen demand for maintenance (suchas elevated temperarure) reduces the maximum size that fish can reach. (C) The relative metabolic rate atfirst marurity (Qm) is necessarily higher than that associated with maximum size (Q-). (D) An evolved,near constancy of the ratio Q../Q- (about 1.4 from guppy to luna) ensures that fish destined to remainsmall (as in case B) also spawn at smaller sizes.

bility to various predators, mainly by their abilityto grow quickly through "small-size" stages in whichmortality is highest. Fish capable of reaching large sizeand that have a high longevity also have low ratesof natural mortality (Fig. 5). Hence fishing tends tohave a stronger impact on species with low naturalmortality, such as sharks or rockfishes. Because theseare often the top predators, their reduction tendsto disrupt the food webs in which they are em-bedded.

C. Relationships between Growthand Mortality

Whichever strategy and tactic fish use to grow, moretime will be needed in large species than in small fishfor the size at first reproduction to be reached. Largesizes thus imply, other things being equal more timeduring which the growing fish may become the preyof some predator. Hence the evolution of large fishwas coupled with a reduction of their relative vulnera-

806 FISH STOCKS

6

5';:~0

N. 4';:0.-.E! 3

2

...cmE00

c{Z0

00 1 2 3 4 5 6 7 8

Aspect ratio of caudal fin

FIGURE 4 Relationship between DNA contents of body cells (a measure of cell size) versuscaudal fin aspect in fish. Note triangular patterns, indicating that active fish with high aspectratios are limited to small cells (which are metabolically more active than large cells), whereasmore sluggish fish may have either small or large cells. Based on records in FishBase 98.

III. DISTRIBUTION OF EXPLOITEDFISH STOCKS

river loaches, Balitoridae) to the mouths of temperateand tropical rivers (e.g., many gray mullets, Mugilidae).In the marine realm, fish range from the intertidal tothe ocean's abyss, both as predators in their desert-likeexpanses (e.g., skipjack tuna, Katsuwonus pelamis) oras components of the rich, newly discovered deep-seavent ecosystems (e.g., some live-bearing brotulas, By-thitidae). Environmental adaptations include the abilityto deal with an enormous range of pressures (from

A. Overall Distribution Ranges

Although mostly confined to water, fish occur in awider range of habitats than any other vertebrate orinvertebrate group. Thus, fish range from the upperreaches of streams in high mountain ranges (e.g., many

FISH STOCKS 807

about one to hundreds of atmospheres), temperatures(from -1.8°C in polar waters to about 40°C in hotsprings, tolerated by some tilapias), and salinities (fromclose to distilled water preferred by the discus fish,Symphysodon discus, of Amazonia to about 10%, e.g.,in West African hypersaline coastal lagoons inhabitedby the blackchin tilapia, Sarotherodon melanotheron),to list only three environmental factors. No single fishspecies or family, however, spans more than small frac-tions of these ranges. Rather, these various adaptationsare exhibited by a bewildering variety of fonns, rangingfrom minute gobies that are fully grown at close to 1cm (e.g., Mystichthys luzonensis) to the 15 m reachedby whale sharks (Rhincodon typus). These two species,incidentally, are exploited for food in the Philippines.The fonner, despite its turnover rate, is in danger ofextinction in the small lake where it is endemic becauseof overfishing and pollution. The latter will be extir-pated if the new directed, export-oriented fishery forthis slow-growing fish continues.

are extracted. Shelves may have rocky or soft (sandyor muddy) substrates, and usually support two weaklyconnected fish communities, one species-rich and con-sisting of bottom or "demersal" fishes, the other con-sisting of fewer species of open-water or "pelagic" fishes.The fish of demersal communities are those exhibitingthe specialized fins and mouths mentioned earlier, en-abling utilization of distinctive food sources, particu-larly on reefs in both temperate and tropical regions.

On coral reefs, this fine partitioning of resourcesculminates in hundreds of fish species sharing a singlereef, with dozens of specialists for each of its foodresource types, from the filamentous algae consumed,for example, by damselfishes (Pomacentridae), the en-crusting algae consumed by parrot fishes (Scaridae),the coral themselves, consumed by butterfly fishes(Chaetodontidae), to the small invertebrates consumedby, for example, wrasses (labridae). A vast array ofpredators such as groupers (Serranidae) and sharks(Carcharhinidae) regulate the number of these smallerfishes. Hard-bottom shelves and, in tropical areas, thecoral reefs that occur down to 30 m are also exploitedwherever they occur. The fishing gear used over hardbottoms are mainly traps and handlines (the latter bothsport and commercial), which are rather selective gearsthat would have relatively minor impacts were it notfor their excessive numbers.

B. Adaptations to Open-Ocean Habitats

Fish have different strategies to deal with the low pro-duction of the oceans. Tuna have adopted a high-energystrategy, wherein their tightly packed schools quicklymove from one food patch to the other, essentiallyhopping from one "oasis" to the next and minimizingthe time spent in the intervening desert-like expanses.Others, notably the lantern fishes (Myctophidae), occurin scattered populations that, at dawn, migrate from1000 m down to the surface waters, and back again atdusk. These different strategies imply very differentbiomasses: tens of millions of metric tons for the majortuna species (prior to their recent depletion by variouslongline, purse seine, and other fisheries) against anestimated global biomass of one billion metric tons forthe lantern fish and associated communities. The latternumber is often viewed as a promising figure, fromwhich various estimates of potential yields have beenderived. Most of these estimates, however, do not con-sider the extremely dilute nature of this biomass (usu-ally less than 1 g per metric ton of water).

2. Demersal Fish StocksThe demersal fish living in, on, or just above shelf softbottoms consist of specialized flatfishes and rays andnumerous generalized teleosts feeding on bottom inver-tebrates (the zoobenthos) and smaller fishes. The com-plex communities thus formed can reach very highbiomass, at shallow depth in the tropics (20-50 m) anddeeper in colder waters. In the warm waters of thetropics, bacteria induce a quick remineralization of thedead organic matter (detritus) falling out of the lightedpart of the water column. This allows very little detritusto become available for consumption by the zooben-thos. In cold water, on the other hand, the short butintensive burst of algal production occurring in thespring is consumed only partly by the zooplanktonof the upper water layers. Most of the remainder isconsumed as detritus after falling down to the sea bot-tom as "marine snow." Thus, cold-water soft-bottom.communities can occur in very deep waters, down tothe shelf slopes (200-300 m) and well beyond. Indeed,the latest trend in fisheries "development" is the exploi-tation of deep-sea stocks of cod-like fish (order Gadi-formes), orange roughy (Hoplostethus atlanticus), andother fish, down to depths of 1000 m or more, through

C. Shelf Communities

1. Definition of Neritic StocksMost fish stocks are neritic, that is, occur above thecontinental shelves, the productive areas of shallowwaters (down to 200 m) around the continents, fromwhich about 90% of the world marine fisheries catches

808 I FISH STOCKS

events, and their subsequent rebuilding, mainly fromrecruits produced off northern Chile.

Pelagic fish tend to form tightly structured, denseschools, which protects them from predators and facili-tates detection and herding of scattered food patches.The fisheries rely on this behavior when deployingpurse seines, which can surround and catch suchschools in one go, often with associated predators suchas dolphins. Large pelagics such as billfish (Xiphiidaeand Istiophoridae) are caught by arrays of longlines, setby the thousands along shelf edges, which also capture,besides the target species, large amounts of by-catch(notably sharks). These sharks were previously left onthe spot, but are now finned before the carcasses arediscarded. Longlines are indeed as unselective as thenow banned giant driftnets that, in the 1980s, erected"walls of death" that were hundreds of kilometers longacross the migratory routes of fish in the North Pacificand the Atlantic.

4. Overall Status of Neritic Stocks

When combined, the demersal and pelagic fisheries ofshelves and adjacent waters represent major threats tofish biodiversity. Particularly endangered are groupersand other slow-growing bottomfish, and pelagics suchas bluefin tuna and various species of sharks and billfish.

Besides the fisheries, one factor contributing to thisendangerment is the traditional separation of researchdevoted to fisheries management ("stock assessments")from that devoted to conservation and to ecosystemresearch. Both lines of research are separated institu-tionally, in terms of their methods and publication out-lets, and in terms of what they perceive as their man-dates. Overcoming this separation is crucial if fishbiodiversity is to be maintained in the face of the on-slaught by fisheries. Key needs are the development oftools and concepts for integrating information on fishbiodiversity and ecosystem function with the knowl-edge gained through a century of applied, single-speciesfisheries research. Before considering these, however,evidence for fisheries impacts on ecosystems will be pre-sented.

ventures that even in principle could never be managedso as to achieve sustainability.

Wherever they occur, soft-bottom shelves are nowa-days invariably subjected to bottom trawling, a veryunselective fishing method that is environmentallydamaging. This involves dragging a heavy, chain-stud-ded net over the sea bottom and "catching," that is,removing all that it encounters. Not surprisingly, thisprocedure has often been compared to harvesting cropswith a bulldozer. Trawler catches thus consist of tar-geted species (usually shrimps in the tropics and sub-tropics) plus a vast number of nontarget species, oftenthe juveniles of demersals with large adult siZes, andliterally parts of the habitat of bottom-fishes, notablysessile invertebrates and chunks of reefs lifted fromthe sea bottom. Nontarget species and debris are thendiscarded, and it is therefore trawlers that contributemost to the global discarding problem. Presently, about30 million metric tons of various fish species are dis-carded; this is a very high discard rate when comparedto the 90 million metric tons that appear in globallanding statistics.

The contribution of trawlers to habitat destruction,including conversion of richly structured bottom habi-tats into featureless expanses of mud, is well recognized,and can only be compared in terms of scale with globaldeforestation and the ensuing trend toward desertifica-tion. Only recently has the impact on biodiversity ofthis mode of fishing begun to be evaluated in systematicfashion. The information so far available indicates highimpacts and a tendency for small generaliZed fish andinvertebrates to replace larger specialized fish, a trendthat amplifies the food web effects to be described later.

3. Pelagic Fish StocksThe pelagic communities over most shelf areas pre-viously consisted of both major and minor stocks andstocklets of herrings, sardines (Clupeidae), anchovies(Engraulidae), and their relatives, and of their preda-tors, notably mackerels and tunas (Scombridae) andvarious jacks (Carangidae). In many parts of the world,pelagic fisheries have eliminated the minor stocks andstocklets, and now depend wholly on annual recruit-ment to the remaining major stocks. The overfishingof old, highly fecund adults in these remaining stocksexplains much of their volatility. Indeed, the presentemphasis of much fisheries research on "variability" isthus devoted largely to a secondary phenomenon cre-ated by the fishery itself. It is true, however, that pelagicstocks, feeding lower in the food web, often closelytrack environmental changes, such as the decline ofthe Peruvian anchovy Engraulis ringens during EI Nino

IV. ECOSYSTEM IMPACTSOF FISHERIES

A. Historical Trends

The earliest fishing gear so far identified by archeolo-gists are bone harpoons that were recovered, along withother evidence of systematic fishing, from a site 90,000

FISH STOCKS 809

notion of sustainable fishing to establish itself. Thisnotion implies that some appropriate level of fishingeffort (number of vessels or gear, mesh size) existssuch that catches (or "yield") can be maintained at highlevels-hence the concept of "maximum sustainableyield" or MSY. This led to the emergence of "fish popula-tion dynamics" and "stock assessments," wherein math-ematical models of single-species fish stocks and of theirresponse to targeted fishing became the mainstay offisheries research. R. J. Beverton, S. J. Holt, and J. A.Gulland in England, W. E. Ricker in Canada, andW. E. Schaefer in the United States proposed most ofthese still-used models during an extremely creativeperiod lasting from the early 1950s to the mid-1970s.

Yet in spite of these advances, the fisheries neverbecame sustainable. One obvious reason was that, givena resource to which access was essentially open, thefisheries never could limit their collective effort at thelevel supposed to generate MSY. Rather, effort levelsincreased well beyond that, permitting some fleet own-ers to increase their stakes even as the aggregate "rent"from the fisheries declined. Recent trends toward subsi-dization of offshore and distant water fleets, driven byinternational competition, have aggravated these eco-nomic issues, enabling commercial profits to be gainedeven from strongly overexploited stocks. These devel-opments are so widespread that they have renderedobvious the impacts which fisheries have on eco-

systems.

years old, in the present~day Democratic Republic ofCongo (formerly Zaire). Tellingly, the main species thatwas targeted appears to have been a now extinct, verylarge freshwater catfish.

This pattern of fisheries exterminating the stocksupon which they ori~nally relied, then moving on toother species, is now understood to be common. Thiscontradicts earlier perceptions of the ocean's quasi-in-exhaustible resources, as expressed among others bysuch Victorian grandees as the geologist Charles Lyelland the zoologist Thomas Huxley. They were misledby the then prevailing abundance of various stocks ofcoastal fish (notably herring, Clupea harengus), andby what may be called "Lamarck's Fallacy": the notionthat "animals living in the waters, especially in sea-water. ..are protected from the destruction of theirspecies by Man. Their multiplication is so rapid andtheir means of evading pursuit or traps are so great thatthere is no likelihood of his being able to destroy theentire species in any of these animals."

The industrialization of the fisheries, first in NorthernEurope and then in North America at the end of the nine-teenth century, quickly showed these predictions to bewrong. Most coastal stocklets of herring and other smallpela~cs were extirpated, and faded even from memory,therein soon followed, after the introduction of bottomtrawling, by coastal stocks of demersal fishes.

The practical response to this was the introductionof bigger boats with bigger engines, fishing farther off-shore. Another response was the creation of researchbodies (such as the International Council for the Explo-ration of the Sea, founded in 1902) to assess the reasonwhy the resources were declining. Also, several coun-tries (notably Norway and the United States) initiatedcostly programs wherein juvenile cod and other fishwere raised in hatcheries and then thrown into the sea,in the vain hope that they would replenish the stocksrather than be eaten by happy predators (which theywere).

B. Emergence of the

Sustainability ConceptThe First World War put an end to the stocking pro-grams. It also established that a strong reduction offishing effort, as caused by the drafting of fishers andvessels into the war effort, and the spiking of majorfishing grounds by underwater mines (thus creating thefirst marine protected areas), would lead to a recoveryof depleted fish stocks. Yet the Second World War, andanother demonstration of stocks rebuilding themselveswhen subjected to less fishing, was required for the

C. Fishing Down Marine Food Webs

The ecosystem impacts of fisheries are due mainly tothe fact that the targeted fish function as part of foodwebs, both as consumers and as prey. Within food webs,the fish of different species occupy distinct trophic lev-els (TL), each defining a step away from plants, whichhave a definitional TL of 1. Thus, fish feeding on plank-tonic algae have TL ~ 2, fish feeding on herbivorouszooplankton have TL ~ 3, and so on. It is important

here to Tecognize that most fish tend to have intermedi-ate TL values (2. 7,3.5,4.1, etc.), reflecting the catholicnature of their diet.

Fisheries, by removing biomass from one of severalfish stocks, necessarily modify food webs, thus forcingpredators of the targeted species to shift toward avail-able alternative prey, if any. Such adjustments werepreviously not distinguishable from natural fluctua-tions. They have gradually become highly visible, how-ever, because they change the mean trophic level of thelandings extracted from different stocks. Moreover, thechanges induced by fishing are not of a random nature,

810 I FISH STOCKS

is true for objective reasons (ecosystems are complex,and their behavior under exploitation, due to the largenumber of stocks to be considered, is difficult to simu-late) and for subjective reasons (notably a perceivedlack of suitable field data on these many stocks).

The recent development of robust ecosystem simula-tion tools should allow the first of these issues to beaddressed. Overcoming the second not only involvespointing out the existence of suitable data, often lostin the "gray literature," but in making such data avail-able in suitable format to all who are aware of the needfor a transition from single-species to ecosystem-basedfisheries assessments. This brings us to the issues relatedto the standardization, dissemination, and uses of bio-diversity information.

V. MANAGING FISHBIODIVERSITY INFORMATION

with decreases in one area matched by increases inanother. Rather, they are directed, with a clear down-ward trend (Fig. 6A), due to the link between growthand natural mortality mentioned in Section II. Thus,in large fish, even a low level of fishing monality gener-ated by a well-managed fishery will quickly exceed thelow level of total mortality (ie., natural + fishing mor-tality) that can be accommodated by the stock. By-catchspecies are even more endangered because the fishingwill not stop as their numbers dwindle until they areeradicated, as has happened with rays in the Irish Sea.The trend of mean trophic level resulting from this(see Fig. 6A), reflecting a phenomenon now knownas "fishing down marine food webs," provides a clearindication that, globally, fisheries generate levels of ef-fort well past those required for sustainability, howeverdefined. Indeed, other indices can be used to indicatethat global changes have occurred in the compositionof global fisheries landings, and in the structure of theecosystems from which these landings are extracted(Fig.6B).

Fisheries-induced modification of the structure ofmarine and freshwater ecosystems has strong indirectimpacts on fish biodiversity, in addition to the directimpacts of reducing the biomass of the target and associ-ated stocks by a factor of 10 or more, as is usually thecase. Incorporating these indirect effects in fisheriesstock assessments has proven to be difficult so far. This

A. Biodiversity as a Conceptual Challenge

3.81

""<D>CD

"0 3.6:ca.e~ 3.4IUCD

:23.2 L-

A.. ...

~

. ".

....

'"100 :5

C>c~80 ~...

~~E E'x 0IG~

.~...

60 E"6c

40 m~I I I I I I I I I I

1950 1955 1960 1965 1970 1975 1980 1985 1990 1995

Year

FIGURE 6 Trend, for the Northeast Atlantic from 1950 to 1996, of two indices of sustained fishing, basedon landings originally compiled by the Food and Agriculture Organization of the United Nations, andother data in FishBase 98. (A) Trend in the mean trophic level of fisheries landings. (B) Trend in the meanmaximum size of fish species in the landings. Note parallel declines, indicative of structural changes inthe ecosystem from which the landings are extracted. Similar trends occur throughout the marine andfreshwater fisheries of the world.

There is a widespread perception that the main obstacleto the conservation of fish stocks and of fish biodiversityis "lack of data," a notion strengthened by public state-ments of biologists worried about the lack of fundingfor relevant research. However, simple lack of data can-not be the problem, not after the 250 years since Lin-

I FISH STOCKS 811

naeus created the taxonomic standards required for bio-diversity research, 100 years of applied fisheriesresearch, and at least 50 years of advances in ecosystemresearch. Rather, the problem here is the fragmentationof the database collected so far. Indeed, many studiesconducted in recent years on the status of various stocksfail to consider previous knowledge on their relativeabundance and distribution, and thus contribute toshifting baselines, wherein only the most recent andusually low estimates are used as reference for conserva-tion or rebuilding efforts.

One reason for this reluctance of biologists to consol-idate existing data into comprehensive, global databasesmay be due in part to the perception that biologicaldata are too difficult to standardize, or are useless oncestandardized. Addressing these issues will be a key taskof biodiversity research, and we now present a few ideasrelated to this.

There is consensus that the objects of biodiversityresearch are genes, populations, species, and ecosys-tems. However, there is little consensus as to whatdistinguishes biodiversity from the existing disciplinesof fisheries biology, ecology, biogeography, populationgenetics, or taxonomy. As a result, the array of databeing claimed to be essential for biodiversity studiesreads like a composite list of the data traditionally usedin the older disciplines, with few attempts at integrationor prioritization. Such integration and prioritization arepossible, however, by giving emphasis, in biodiversitystudies, to data that are: (1) relevant to current researchissues (e.g., richness, rarity, distinctiveness, representa-tiveness, threat, function, and utility of species); (2)part of the data traditionally collected in taxonomy,biogeography, population genetics, and ecology; (3)widely available, in sufficient quantity; (4) pertinent topast, present, and most likely future trends; (5) easy tocollect; (6) easy to standardize; (7) easy to verify; and(8) suggestive of new lines of research.

richness (number of species encountered) is deriveddirectly from the bioquads from a given area. Distinc-tiveness (how much the species encountered differ fromeach other) is derived from the classification of thesespecies into higher taxa such as families, orders, andclasses. Representativeness (how closely an area repre-sents a predefined ecosystem type) is derived by com-paring observed species composition with the typicalcomposition of the ecosystem type under study. Theutility of species to humans can be derived from pub-lished, or local knowledge, or from catches in the caseof fish. Status of threat can be derived from trends inthe distribution area defined by bioquads. Rarity canbe estimated from the number of bioquads available fora species in a given area, standardized by samplingeffort.

Taxonomists have made a conscious effort to system-atically compile data of this sort in specimen collections,and to publish them in original species descriptionsand revisions. As a result, bioquad-type data are readilyavailable in enormous numbers (about 10 million forfish alone) in museum collections, survey reports, his-torical photos and films, and other forms (criterion 3).While museum collections go back over 200 years, someliterature contains verifiable records that date back toantiquity (criterion 4). Also, archeological data reachback to the dawn of modem humanity (see the earlierrecord pertaining to giant catfish).

Numerous scientific surveys and projects also con-tinuously collect contemporary bioquads. Othersources are the commercial fisheries and the many lay-persons whose hobby is to observe and sometimes tocollect fish and other wildlife. These activities are mostlikely to continue in the foreseeable future (criterion5). An increasing number of the preceding data sourcesare available in computer-readable form (criteria 3, 5,and 6).

Efforts do exist to standardize the elements of thebioquad (criterion 6). For example, the Species 2000Initiative has embarked on the task of providing a stan-dard reference list of the valid names of the known 1.75million species sharing Earth with humans (see thewebsite www.sp2000.org). Geographical coordinatesand the international date and time format are obviousstandards for items (2) and (3), although there remainsa need for a global gazetteer to deal efficiently withlocalities reported without coordinates, and there is aneed for standards to deal with date and time ranges.On the other hand, standards exist for sources suchas printed publications, databases, photos, films, andpersonal communications. Many of these were consid-ered when developing FishBase, a computerized data-

B. Bioquads as Key BiodiversityData Sets

A minimum core of biodiversity information that fulfillsthese eight criteria is provided by "bioquads" (from"quads," short for quadriads), consisting of: (a) thescientific name of a taxon, usually a biological speciesor other evolutionarily significant unit; (b) the localitywhere a specimen of this taxon has been encountered;(c) the date (time) of the encounter; and (d) the author-ity or source reporting (a)-(c).

Of the research items mentioned under criterion (1),

812 .' I FISH STOCKS

base on the biology ,ecology, and uses of fish containinga vast number of bioquads (see the following).

The necessary verification (criterion 7) of millionsof data points can only be done automatically. Basically,a computer can verify a scientific name against a stan-dard list, compare the indicated locality and date againstthe established range of a species, and judge the reliabil-ity of a source, for example, by the number of outliersit has reported prevIously. Procedures will have to beestablished, however, on how to deal with the differenttypes of outliers, some of which may represent validnew information.

An important consideration is how fast a researchagenda based on bioquads will be exhausted (criterion8). Important here is the ability of well-structured rela-tional databases to interlink independently developeddata sets. Thus, the scientific name links to all availableinformation on a species, including taxonomy, system-atics, genetics, biology, ecology, and human uses. Thelocality connects to all available information on sur-rounding environments, including province, country,continent, habitat, ecosystem, and tectonic plate. Thecombination of species, locality, and date points to apopulation or stock. Date and time in connection withthe locality can be used to infer a wide range of environ-mental conditions, from local temperatures to currentfisheries legislation. The source relates to the humandimension, such as persons and institutions workingon certain species groups or in a certain area, represent-ing the scientific interface between humans and theother species (Fig. 7).

FIGURE 7 Interrelationships of the elements of biodiversity, articu-lated through the four elements of bioquads (species, location, times,and source).

bioquads as defined previously. Important here is thata new original of this graph is generated on the fly, fromcurrently available bioquads, every time the relevantroutine of FishBase is evoked, and that each of its "dots"can be clicked to verify the four elements of the underly-

ing bioquad.

VI. PRESERVING FISH BIODIVERSITY

A. Traditional Approaches to

Stock Management

None of the foregoing considerations will help, how-ever, if fisheries are allowed to continue underminingtheir resource base, which they will if fisheries manage-ment continues to rely on the panoply of approachesso far deployed. These traditional approaches include,among other things: (1) mesh size restriction; (2) re-striction on the amount and/or species of fish that maybe legally landed; (3) effort limitation, for example,through caps on the vessel tonnage that may deployed;and (4) seasonal closures.

Besides being extremely hard to enforce, these ap-proaches-which are invariably conceived in the con-text of single-species assessments-fail to address theecosystem effects mentioned earlier. Thus, mesh sizes

C. Databases as Tools for Management ofBiodiversity Information

Two major initiatives presently exist to assemble andmake widely available, for research on fish biodiversity,the information presendy held by various institutions(notably museums). One is NEODAT, which makesaccessible on the Internet about 400,000 bioquad rec-ords pertaining to freshwater fish of the Neouopics(NEODAT; www.fowler.acnatsciorg). The other isFishBase, an ongoing international collaborative projectdedicated to assembling the estimated 10 million ex-isting fish bioquads and to combining them with other,standardized biological information on fish. The inten-tion here is to provide a global relational database,addressing head-on the data fragmentation issue men-tioned earlier (see www.fishbase.org).

Figure 8 shows the geographic distribution of Niletilapia, Oreochromis niloticus, through dots representing

FISH STOCKS 813

.

FIGURE 8 Distribution of Nile tilapia (Oreochromis nilotiC1tS) based on 425 bioquads contributed by the Musee Royal de l'MriqueCentrale, Tervuren, Belgium, and other sources. In the computerized version of this graph, each dot can be "clicked" to revealthe four elements of the underlying bioquad, thus allowing identification of outliers, temporal trends, etc.

above a certain limit, meant to protect the young of agiven species, do not prevent associated species formbeing caught. Indeed, when combined with restrictionson total allowable catch (TAC), and on the landing ofbycatch (as is often the case), mesh size restrictionsbecome the very reason for discarding both the youngof targeted species and the nontarget species. Limits onnominal fishing effort are subverted by technologicaldevelopments, such as improved gears and navigationinstruments (e.g., GPS), which increase the catchingpower of fishing vessels. Thus, government-run vesselretirement schemes often end up subsidizing the mod-ernization of fishing fleets. Finally, seasonal closure ofvarious areas usually has negligible ecological impacts,because the fishing effort expended during the openseason is sufficient for the sea bottom to be scraped upnumerous times by trawlers, and for the stocks of long-lived fishes to be severely impacted.

B. ~arine Protected i\reas

There is an emerging consensus among fisheries scien-tists and conservationists that the only fisheries man-agement tool that will allow the recovery of damagedstock and ecosystems is the establishment of MarineProtected Areas (MP As), including permanent N 0- Takezones as their core. Such core zones are easy to en-force-at least relative to the task of enforcing meshsizes or TACs. Also, technology-driven increases offishing effort can be ignored, and there is assurancethat the long-lived organisms of seafloors and theirassociated fish communities can gradually return to asemblance of their original configurations. However,much research will have to be devoted to identifyingthe optimal size and location of MP As, particularly formigratory stocks.

Still, traditional fisheries management, aimed at lim-

814 FISH STOCKS

iting effective fu;hing effort, will have to continuearound MP As, lest they become marine larders or fu;h-attracting rather than fu;h-producing zones from whichresources are drained by fu;heries operating at theirvery periphery.

Finally, the social context of fu;heries will have tochange: fu;heries do not harvest crops they have sown.Rather, they exploit the natural productivity of wildlife;thus there are inherent limits to global fu;h catches, andfuture fu;heries will not meet the demand of an ever-increasing human population. Indeed, the massive eco-system changes already described indicate that theselimits have been reached in most parts of the world,and that sustainable fisheries must be embedded insome form of ecosystem management.

See Also the Following Articles

Bibliography

ADAPTATION. FISH, BIODIVERSITY OF. FISHCONSERVATION. MARINE ECOSYSTEMS

Froese, R., and D. Pauly (eds.). (1998). FishBase 98: Concepts, Designand Data Sources. ICLARM, Manila. [Distributed with two CD-ROMs; also see the website www.fishbase.org.]

Hawksworth, D. L, P. M. Kirk, and S. D. Clarke (eds.). (1997).Biodiversity Information: Needs and Options. CAB International,Wallingford, United Kingdom.

Miller, R I. (ed.). (1994). Mapping the Diversity of Nature. Chapman &:Hall, London.

Mooney, P. (ed.). (1998). Ecosystem management for sustainablefisheries. Ecol. Appl., Supplement to 8(1).

National Research Council. (1999). Sustaining Marine Fisheries. Na-tional Academy Press, Washington, D.C.

Nelson, J. (1994). Fishes of the World, 3rd ed. John Wiley &: Sons,New York.

Pauly, D. (1994). On the Sex of Fishes and the Gt:nd£T of Scientists:Essays in Fisheries Science. Chapman &: Hall, London.

Paxton, J. R., and W. N. Eschmeyer (eds.). (1998). Encyclopedia ofFishes. Academic Press, San Diego.

Reaka-Kudla, M. L., D. E. Wilson, and E. O. Wilson (eds.). BiodiversityII. Understanding and Protecting our Biological Resources. JosephHenry Press, Washington, D.C.