Introduction

Naturally derived fish production from inland and inshore marine systems is becoming increasingly limited for two reasons. Firstly native fish assemblages have been unable to adapt to the declining quality of the aquatic environment and secondly, many fish species fail to compensate for excessive fishing through natural reproduction. As a result of these two stresses catch from inland fisheries based on naturally reproducing fish populations is declining and catches in some coastal areas are likewise threatened. One response to this crisis has been to use stocking with selected species to mitigate the damage and to maintain or increase catches. In this way some conventional inland and coastal fisheries are being transformed from capture fisheries to a form of aquaculture.

There are diverging scenarios for the management of inland and coastal waters for fisheries based mainly on differences in social, cultural and economic factors. These differing views, which are simplified in Table 1, have implications for the ways in which the waters are managed and, in particular, for the development of stocking programmes and of aquaculture.

In developed countries stocking coupled with habitat maintenance became the major form of inland fisheries management relatively early in the present century. More recently, aquatic resource use is becoming increasingly subordinated to conservation. Production facilities are generally isolated in controlled fish-farms and the inland waters are destined mainly for aesthetic and recreational uses. Ideally in this regime stocking programmes are limited to low numbers of large fish in support of native species or restoration of endangered ones, but exceptions are common in lakes where capture fisheries are supported by large scale stockings and in some recreational fisheries which are still intensive and are maintained by high stocking rates with smaller fish. In the sea the collapse of many inshore grounfish resources is inducing some countries to compensate by stocking finfish, crustaceans and molluscs.

In developing countries there is a large food deficit, and inland fisheries are called on to maximize yields. As a consequence most inland waters reached their maximum potential under natural production some time ago, and rising demand is now pushing many tropical waters to maximize yields through enhancement. In many countries this process is now advanced with the development of infrastructure for the production of fingerlings for stocking. In other areas lack of funding and infrastructure is delaying the process despite there being adequate physical potential. In many cases there is a close integration between capture and culture through extensive and semi-intensive management of artificial water bodies and rice paddies.

Stocking programmes in both regimes have been pursued uncritically with little scientific evaluation of their success or failure. Whilst the primary motive for stocking was political, either to convince recreational fishermen that Government authorities were looking after their interests or to subsidize fishermen's communities which would otherwise have insufficient catch, the need for evaluation was not pressing. However, where stocking is used as a tool to maximize yields, outlay in stocking material is usually the major cost and thus a more rational approach to the use of resources can result in considerable savings.

Major stocking programmes

Although 94 countries have reported stocking to FAO as part of their fishery statistics only a few countries have documented the success or failure of their programmes. Examination of the scientific literature produces some 4, 847 references extracted from ASFA (1978-96) dealing with stocking of finfish, molluscs or crustaceans into inland or marine habitats most of which do not help understand the rationale of the stocking exercises. Most documented evaluation of stocking programmes appears in national or international reports and is drawn from very few countries including China, Cuba, India, Mexico, Poland, Russia and Sri Lanka for inland waters and Canada, Iran, Japan, Norway and the U.S in marine fisheries.

Stocking with major carps has been the main policy for management of Indian reservoirs since the early 1960s. Programmes were initially unregulated, with no evaluation of their effectiveness, but where a more scientific approach has been adopted, yields and profitability have increased substantially. For example effective stocking policies raised production from 1.67 kg/ha in 1964 to 194 kg/ha in 1985 in one reservoir, whilst in another yields rose from 10 kg/ha in 1966 to 107 kg/ha in 1991. Total and the percentage of commercially valuable major carps in the catch were substantially raised in many other reservoirs (Sugunan, 1995).

In other parts of Southeast Asia (Thailand, Philippines, Malaysia, Indonesia and Sri Lanka) where natural productivity is about 20 kg/ha (Fernando, 1977) stocking in inland waters raised yields to over 100 kg/ha. In China, the very intensive stocking rates of 2,250-4,500 fish/ha of fertilised reservoirs have substantially increased in yields and profitability.

Reservoirs of the former USSR (Berka, 1989) have been managed by stocking since the late 1960s. Stocking rates and procedures were defined for various species of fish and accompanied by measures to improve fertility, the structure of the lake and the elimination of the pre-existing fauna. Yields were considered far higher in managed lakes at about 250 - 500 kg/ha than those left to natural processes.

Stocking is not universally effective however. Systematic studies of 60 Cuban reservoirs for up to 13 years (Fonticiella et al., 1995) showed that for species that breed naturally and for which adequate spawning substrates exist, stocking exercised little influence on fish catch. Sugunan (1995) also suggests that stocking has proved ineffective in some Indian reservoirs.

Origin of stocking material

Seed material for stocking extraction either comes from natural spawning from rivers and lakes or from aquaculture installations.

Natural production

Early stocking and aquaculture were based heavily on seed from natural sources and in several species, such as shrimp, milkfish, yellow tail and eel, stocking continues to be based on natural production today. The practice of extracting large numbers of young fish from rivers and lakes has caused concern to fisheries managers on the basis that this would damage the sustainability of stocks. Welcomme and Hagborg (1977) proposed that under normal conditions a considerable proportion of the young-of-the-year can be removed from floodplain rivers due to the overproduction of 0+ fish from such fluctuating systems. The situation in lakes and regulated rivers would appear more precarious due to the lesser flexibility in the dynamics of stocks from more stable systems. The sustained withdrawal of large numbers of fry from Chinese and Indian rivers over a period of several decades would appear to support the contention that river stocks can be used as a source of supply for stocking material. However, a combination of overfishing and environmental degradation have had impacts on the stocks. For instance, in Chinese rivers, fry availability and fish populations have declined substantially mainly due to the exploitation of the stock (Lu, 1994). The Ganges river provided 90% of the seed needed for carp aquaculture in 780000 ha of ponds and tanks in India in 1964, and by 1989 these numbered 3.7 billion individuals. At these levels of offtake the fisheries for the major carps were judged to be collapsing with a seed supply that was reduced both in quantity and quality. At the same time the percentage of major carps in the catch and in seed supplies was declining in favour of minor carps (Natarajan, 1989). In 1995, seed collection apparently still formed a major industry in the middle and lower Ganges.

Hatchery production

While most early stocking was carried out with material of natural origin, a few species which could be induced to spawn through simulated natural processes were stocked from hatcheries even in the last century. Indeed, many of the early salmonid hatcheries in Europe and North America dating from the end of the last century were set up for stocking of sport fisheries rather than for the production of table fish. Early hatchery production concentrated on producing massive amounts of eggs or larvae, but as size at release is now realised to be an important factor in the success of stocking, release generally takes place at a more advanced stage. The development and spread of techniques for controlled, induced spawning in the 1950s and 1960s generally freed the aquaculture sector from natural sources of supply and at the same time provided an alternative source of seed for stocking a greater range of species.

Size of fish stocked

Two main factors influence the size chosen for stocking material, cost and survival. The penalty of high mortality acts against the stocking of fish at too small a size, while the increased cost of the stocking material with increasing size tends to favour stocking of early life stages, especially in slow growing species. The actual size chosen usually depends on an empirically determined balance between these two factors as well as on the life history of the fishes. Migratory and anadromous fishes such as salmonids are usually stocked at a small size (fry) to prepare for migration as their size increases, whereas cyprinids are generally stocked at a larger size (fingerlings).

Dynamics of stocking

The dynamics of stocking are apt to be particularly complex when fish are stocked into systems where there is also natural reproduction of the stocked species. Mortality may increase and growth rates decrease due to the addition of excess elements to the stock. Clearly these effects will depend on the relative proportions of stocked fish to those originating from natural reproduction. It is clear that where there is adequate natural reproduction stocking may well be superfluous, as in the case of tilapias in Cuban reservoirs (Fonticiella et al.,1995). In some fisheries, such as the Finnish coregonid fisheries and the Polish lake fisheries, stocking has made a long term positive contribution despite there being naturally reproducing stocks of the target species (EIFAC, 1983). In other lakes, for example in Thailand (Bhukaswan,1988), it has been impossible to evaluate the biological and economic success of stocking due to difficulties in separating stocked fish from natural production.

Analysis of the few data sets on stocked systems where the stocked species does not reproduce indicates an interplay of two main variables - the area of stocked system and the stocking rates. Thus:

i) Yield is related to stocking rate. Relationships between stocking rate and catch may be obtained for one lake, for example the Nanshahe reservoir in China where Yield in kg/ha is directly proportional to stocking density (Li and Xu, 1995), or for a series of lakes as in the case of Sri Lanka or Mexico. Relationships are usually assumed to be linear but some interpretations, such as that of Amarasinghe (in press), predict curvilinear relationships between stocking rate and yield, with an initial rise in yield but a later fall as density dependent factors come into play.

ii) Yield per unit area is inversely related to the area of the stocked system. Stocking has generally proved more effective in small reservoirs in many regions although larger water bodies are also stocked. Data sets from China (Li and Xu, 1995), Sri Lanka (Amarasinghe, in press) and Mexico (FAO, 1992, 1993) all show strong inverse relationships between reservoir area and yield per unit area which are linear on log-log scales. Fonticiella et al., (1995) found a similar influence of area in Cuba where there was a very strong correlation between stocking density and catch in small reservoirs managed semi-intensively but a lesser correlation in extensively managed larger water bodies. But...

iii) Stocking rate is inversely related to lake area. In nearly all countries practising stocking there is a tendency to stock fish at lower densities into larger water bodies. This is generally justified by a) the supposition that competition and predation will be greater in larger water bodies and therefore survival rates will be reduced, and b) that other aspects of enhancement such as fertilization, control of unwanted species or construction of artificial faunas are more controllable in smaller water bodies. If the first of these suppositions were true then the efficiency of stocking in terms of yield per unit stocked would be expected to fall as the size of the stocked water bodies increased. In none of the cases examined was there a correlation of this type and efficiency of stocking remained roughly the same over a vast range of reservoir areas. It must be concluded therefore that, at present, the practice of using small water bodies for intensive and large ones for extensive stocking depends more on the high number and cost of fingerlings needed for stocking the larger water bodies, and the difficulties of controlling other parameters, including poaching, in the larger reservoirs.

Although stocking in inland waters and that of marine systems are usually treated as separate topics, there is little biological justification for this division. Because of their size marine systems tend to be stocked at low levels relative to the whole area available. However, for species that are relatively sedentary, or those stocked and harvested from nearshore waters before moving out to sea, the stocking density related to the area of residence may be much higher than is apparent.

Summary of stocked system

The increases in productivity and the associated enhancement strategies at various levels of stocking are described in Table 2.

References

Amarasinghe, U.S., In Press. How effective are the stocking strategies for the management of reservoir fisheries of Sri Lanka.

Berka, R., 1990. Inland capture fisheries of the USSR. FAO Fish.Tech.Pap.311.143p.

Bhukaswan, T., 1988. The use of cyprinids in fisheries management of the larger inland water bodies in Thailand in Petr, T. (ed.) Indo-Pacific Fisheries commission, Papers contributed to the Workshop on the Use of Cyprinids in the Fisheries Management of Larger Inland Water Bodies of the Indo-Pacific. FAO Fisheries Report.405:142-150.

European Inland Fisheries Advisory Commission, 1983. Report of the EIFAC Working Party on Stock Enhancement. EIFAC Tec.Pap.44:22p.

FAO, 1993. Avances en el manejo y approvechamiento acuicola de embalses en America Latina y el Caribe. GCP/RLA/102/ITA, 6/95: 18-23.

FAO, 1992. Maejo y explotación acuicola de embalses de agua dulce en America Latina. GCP/RLA/102/ITA. Doc de Campo 8. 162p.

Fernando, C.H., 1977. Reservoir fisheries of Southeast Asia: past, present and future. Proc.IPFC,17(3):475-489.

Fonticiella, D.W., Z. Arboleya and G. Diaz, 1995. La repoblación como forma de manejo de pesquerias en la acuicultura de Cuba. COPESCAL doc.ocasional, 10:45p.

Li S and S. Xu, 1995. Capture and culture of fish in Chinese reservoirs. IDRC, Ottawa: 125p.

Lu, X., 1994. A review of river fisheries of China. FAO Fisheries Circular, 862:47p

Natarajan, A.V., 1989. Environmental impact of Ganga basin development on gene pool and fisheries of the Ganga river system in D.P. Dodge (ed.) Proceedings of the International Large Rivers Symposium, Can.Spec.Publ.Fish.Aquat.Sci., 106:545-560.

Sugunan, V.V. 1995. Reservoir fisheries of India. FAO Fish. Tech. Pap. No. 345: 423p. Rome, Italy.

Welcomme, R.L. and D. Hagborg., 1977. Towards a model of a floodplain fish population and its fishery. Env. Biol. Fish., 2:7-24.