1

E5-05-04-02

FRESHWATER AQUACULTURE AND POLYCULTURE

Lionel Dabbadie and Jerome Lazard Unité de recherche Aquaculture, CIRAD-EMVT, Montpellier, France

Contents

1. Brief Review of the State of World Freshwater Aquaculture 2. The Different Types of Freshwater Fish culture and Polyculture 3. The Main Cultivated Species in Freshwater Aquaculture 3.1 Carps 3.1.1 Grass Carp 3.1.2 Silver Carp 3.1.3 Bighead Carp 3.1.4 Common Carp 3.2 Catla 3.3 Rohu 3.4 Mrigal 3.5 Tilapias 4. Freshwater Fish Farming Management 4.1 Reproduction 4.2 Nutrition and Feeding 4.3 Fish Pond Ecosystem Management 4.3.1 Pond Dynamics 4.3.2 Chemical Fertilization 4.3.3 Organic Fertilization 4.4 Fish Populations Management 4.4.1 Polyculture and Monoculture 4.5 Pond Stocking 5. Freshwater Fish Farming Economics 6. Conclusion and Prospects

Glossary

Amino acid: Organic compound with a carboxyl (-COOH) and an amino (-NH2) group. Autotrophy: Assimilation of mineral compounds to produce organic biomass. C/N: Carbon to nitrogen ratio. Carbohydrates: Organic compounds whose general composition is: C x (H2O) y. Cladocera: Aquatic crustacean belonging to zooplankton. Copepod: Aquatic crustacean belonging to zooplankton. CSC: Critical standing crop, biomass at which fish growth declines. DL50: Quantity of a toxic product that entails the mortality of 50% of a population, after a certain time

(96 hours generally). Eukaryotic: Cell with nucleus. Fatty acid: Organic acids with a methyl (-CH3) and a carboxyl (-COOH) termination, that differ in

the number of carbon atoms in the molecule, and in the number and positioning of the double bonds between the carbon atoms. Fatty acids with one or more double bonds are unsaturated fatty acids. The formula (n-3) denotes that the first double bound of a fatty acid is found in the link between the third and fourth carbon atoms from the methyl end.

FCR: Food Conversion ratio, ratio between the quantity of distributed food and the weight gain of fish.

Fertilization: Spreading of mineral or organic compounds that stimulate the natural pond productivity.

2

Fingerlings: Young fish larger than a fry but not adult. GnRH: Hormones produced by hypothalamus that induce gonadotropin production by hypophysis. Gonadotropin: Hormones produced by hypophysis, which induce sex steroid production by gonad. HCG: Human Chorionic Gonadotropin, gonadotropin hormone used for artificial fish reproduction. Heterotrophy: Opposite of autotrophy. Integrated systems: Farming systems where aquaculture is integrated to other agricultural

productions (rice, breeding etc.). LHRH: Luteinizing-Hormone Releasing Hormone, a human GnRH also used for fish artificial

reproduction. K: Carrying capacity, maximum biomass that can be carried in a pond. Limnology: Science of freshwater. Lipids: Water-insoluble organic compounds. Mineralization: degradation of organic matter into mineral salt and inorganic carbon by bacteria. Nauplius (plural: nauplii): Young stage in some crustacean larval development (used for copepods

in the text). Phytoplankton: Vegetal component of plankton. Plankton: Aquatic organisms suspended in the water column. Polyculture: Cultivation of more than one species with complementary feeding habit and behavior in

a single place. Primary production: Photosynthetic-dependant biomass production. Protein: Large organic molecule that contain carbon, hydrogen, oxygen, and nitrogen and often sulfur,

made of amino acids. Protozoa: Unicellular eukaryotic organisms. Rotifer: Aquatic worm belonging to zooplankton. Sex steroids: Hormones responsible for gamete maturation. Stocking density: Number of fish stocked per unit area of a pond. Zooplankton: Animal component of plankton. Zoug bottle: A classical device for incubating fish eggs.

Summary

Fish is an important food component for populations of many countries, especially in developing areas. For food fish, over a quarter of total world supply is derived from aquaculture and contribution of the latter increases quickly. Several types of fish culture can be distinguished, according to their degree of intensification, or depending on the fish feeding: mostly natural food or exogenous high protein feed. Pros and constraints of each type are reviewed, but pond production remains the most used as it permits a multipurpose production (fry, fingerlings, adults) with a wide spectrum of utilization (from extensive to hyper-intensive). The main cultivated freshwater species belong to carps (Indian and Chinese carps) and tilapia. Their biological characteristics are reviewed. Management and improvement of environmental factors can achieve fish reproduction in captivity, but most frequently, it is done using hormonal treatment: pituitary extracts, gonadotropin or GNRH with or without antagonists-inhibitors.

Fish nutrition requires high levels of protein, but expensive fishmeal can be partially replaced by vegetal protein for some herbivorous species (tilapia), or by lipids and carbohydrates as many proteins are used as an energy source. But the most important factor in many aqua cultural farms is the management of the pond ecosystem and the definition of adequate stocking. Natural feed for fish can be stimulated by fertilization. Mineral fertilizers stimulate the autotrophic-based food chain, whereas organic manures improve both the autotrophic and heterotrophic food chain, thus allowing higher fish crops. Pond stocking must be defined in order to be as close as possible to the bio-technical-economic optimum. When too low, natural productivity is under-exploited and yield is low. When to high, fish growth decreases quickly and use of an expensive artificial feed become necessary. Another way of improving pond productivity consists in breeding several species with complementary feeding regimes. This is polyculture and several examples practiced around the world are given. Finally, some fish farming systems are considered from the economical point of view. Integrated systems have a higher profitability and systems that associate aquaculture to

3

livestock breeding may show a contribution of fish to net income higher than 50%. These trends are encountered in several areas worldwide.

Introduction

Freshwater aquaculture is by far the most ancient aquatic living resource production system known in the world. It strikes root in more than 2500 years of history. Fish is the main, if not the only, component of freshwater aquaculture and earthen pond is historically the first and still the most utilized aquaculture production facility (contribution for more than 80–85% of the total freshwater production).

Freshwater aquaculture differs from other aquaculture systems by some characteristics. It allows a strong integration to the agricultural production systems (crops and livestock) at different levels: water use, wastes recycling into the fishponds as fertilizers, agricultural by-products as fish feed. Freshwater aquaculture production is mainly based on the culture of short food chain fish (carps, tilapias) and differs basically from marine fish culture based on carnivorous fish (salmon, Japanese amberjack). Freshwater aquaculture is mainly based on extensive and semi-intensive aquaculture production systems where polyculture, fertilization, and supplementary are the key points.

In 1995, freshwater aquaculture accounted for 65% of the total aquaculture production, if aquatic plants are excluded. Asian countries are the main aquaculture producers and in these countries, freshwater fish culture plays a major role: China, India and Indonesia (67%, 6.7%, and 3.1% of the total world aquaculture production respectively).

In the last decades, major bio-technical innovations have had a strategic impact on the freshwater aquaculture development: artificial breeding, use of supplementary feeding and artificial feed, genetic improvement, introduction of exotic species to many countries for aquaculture purposes. Despite all the scientific studies carried out in this field, the pond as a culture environment remains a black box where fish feeds at many levels of the food web and fish species interact actively.

The progresses in pond fish culture management practices have mainly been obtained by a trial and error process. This contribution aims at presenting the main available data concerning scientific and technical bases, and practices in the field of freshwater aquaculture.

1. Brief Review of the State of World Freshwater Aquaculture

In 1995, total world production of finfish, crustaceans, mollusks and plant reached 120.7 million mt, of which aquaculture contributed for 27.8 million mt. The annual contribution of aquaculture to total aquatic production increased from 14.4% in 1989 to 23% in 1995. For food fish, over a quarter of total world supply was derived from aquaculture.

Regional, cultural, and historic attributes have played a major role in influencing both the production base and rate of expansion of aquaculture.

The historic tradition of growing fish in Asian countries such as China, India and Indonesia, which are the three leading countries, has played a significant role in maintaining Asia’s dominant position in aquaculture. Aquaculture production in Asia increased from 1984 to 1995 at an average annual compounded growth rate of 10.4% and accounted for over 90% of world output. Much of this growth, however, specifically relates to China. In Africa and Latin America, the aquaculture production base is considerably lower but even with low total productions compared with Asia, the growth rate reached 12.7% in Africa and 12.8% in Latin America during the same period as above, while the growth rate was 3.9% in Europe and 3.6% in North America.

If aquatic plants are excluded from total aquaculture production, the contribution of finfish and shellfish from freshwater environments continues to dominate output and in 1995, accounted for

4

around 63% of the total tonnage of cultured finfish and shellfish. Culture from brackish water and marine environments contributed 7% (mainly crustaceans) and 30% (mainly mollusks) respectively.

Of the 292 species included in the 1995 FAO aquaculture statistics for which production data are available, the first 22 species accounted for 80% of the total production. Of these 22 species, nearly all of the farmed animals are filter feeders, herbivores, or omnivores. Only one species, Atlantic salmon, is carnivorous and is clearly a minor species in terms of production volume.

The most important group is freshwater finfish: 12.7 million mt compare with 1.4 million mt of diadromons fishes and 0.6 million mt of marine fishes.

The freshwater finfish production is dominated by Cyprinids and tilapias with a contribution of 10.3 million mt in 1995, cyprinids have a number of advantages which will maintain their leadership in the short and medium term: they can use feeds with moderate protein and fish meal content. They can be reared in polyculture systems that make optimum use of the natural productivity of the ponds and water bodies where they are stocked. They have also good markets in Asian countries due to traditions and relatively low prices.

Only a few carp species dominate global Cyprinids culture: the Chinese carps (silver, grass, bighead, crucian, black, and mud), common carp, and Indian major carps (rohu, catla, and mrigal). In 1995, these species accounted for 80% of all cultured carps. The culture of Chinese carps was dominated by three species: the silver, grass, and bighead carps (70% of total Chinese carp production in 1995). Common carp made up another 21% and is geographically the most widespread, being cultured in 86 countries. In 1995, the larger producers of common carp were outside China, India, Indonesia, Russian Federation, and Ukraine.

All landings of Indian major carps were from aquaculture and culture of the three main species increased at an annual rate of 12% from 1984 to 1995. Almost all-Indian major carp production has come from India but in recent years a growing proportion is cultured in Myanmar, Thailand and Laos.

Between 1984 and 1995, the contribution of cultured tilapias to total tilapia increased from 38% (198 000 mt) to 57% or 659 000 mt. Four cichlid species or species groups (Nile tilapia, unidentified tilapias, Mozambique tilapia and blue tilapia) dominated production between 1984 and 1995 where they accounted for 99.5% of cichlid production. Nile tilapia accounted for 72% of total tilapia production in 1995 and its annual percent growth rate from 1984 to 1995 was 19%. In 1995, major tilapia producers are China (315 000 mt), Philippines (81 000 million mt), Indonesia (78 000 mt), and Thailand (76 000 mt).

2. The Different Types of Freshwater Fish culture and Polyculture

The different fish farming production systems are generally distinguished according to their degree of intensification which is itself usually defined according to the feeding practices as food represents more than 50% of the total operating costs in intensive systems. However, intensification (or, inversely, extensification) involves many other production factors, such as water, land, capital, and labor.

A first classification could be established as follows:

• Extensive fish farming production systems are based on the use of natural feed produced in the

fish culture structure/environment without any input with a very low input level. Rice-fish farming systems can be considered as belonging to this extensive level as fish takes benefit of inputs added for rice cultivation

• Semi-intensive fish farming systems rely on the use of fertilization (organic and/or mineral) to produce natural feed and/or supplementary feed, but with a significant amount of the fish diet

5

supplied by natural feed. Integrated crop-livestock-fish farming systems are typically belonging to this type of fish culture as well as all fish farming systems recycling various types of wastes including direct excreta reuse systems (latrine ponds of Vietnam for example) and indirect sewerage systems. Both systems provide high fish yields

• Intensive and super intensive systems have all the fish nutritional requirements provided by a nutritionally complete pelleted feed with little or no nutritional benefits from natural productivity of the pond or water body where fish culture is achieved (lake, river). The feed used in these fish farming systems are generally rich in proteins (25 to 40%) and are then costly. The main facilities used for this type of fish farming are pens, cages or raceways with a very high water renewal rate (natural through water currents of artificial through pumping)

The different types of fish farming systems according to the level of intensification are summarized in Table 1.

Table 1. Different levels of intensification of fish farming systems.

An interesting transition between semi-intensive pond fish culture and super-intensive fish culture systems is given by tilapia floating cage culture in productive natural water bodies such as lakes in the Philippines. The stocking rate of tilapia fingerlings is adapted to the cage size, the natural productivity of the water and the culture management. At low stocking densities (up to 25 fish.m-2), supplemental feeding may not be necessary especially in productive lakes during the abundant plankton season. To accelerate fish growth during the low productive months, supplemental feeding is applied at rates of 3 to 5% of body weight per day (see Table 2).

Table 2. Stocking rates for Nile tilapia in cages of different sizes and management schemes.

Another typology of fish farming production systems can be proposed, based only on the discrimination between systems where the feeds originate only (or mostly) from the ecosystem (endogenous feeds produced by the ecosystem) and systems where feeds are entirely exogenous and where fish feeding is entirely based on pelleted feeds or even trash fish.

Management of the first type involves fertilization and/or supplementary feeding, polyculture practices and there is a very strong interaction between stocking rate, final individual weight of fish (growth rate) and yield, which have to be managed very thoroughly. Management of the second type relies mainly on monoculture, high stocking rates and artificial feeding with a high protein feed.

The choice of any of these two types of fish culture systems depends on many factors, which are listed in Table 3.

Table 3. Types of fish culture systems according to the different feeding management practices.

In terms of land requirements, for a given prospected fish production, ponds need much more area of land (or water surface) than more intensive fish culture systems which, on the other hand, need higher water flow rates provided by natural currents in lakes or rivers (in the case of cages or pens) or by water supply by gravity or artificial pumping (tanks, raceways). Ponds used for fish culture have generally a low level of negative impact on the environment. Moreover, they can be used for recycling various types of wastes. For example, ponds can recycle wastes such as night soil directly as sewage-loaded ponds where fish are cultured or indirectly within an aquaculture excreta rinse system with stabilization and maturation ponds, the fish being farmed in the latter.

6

Capital for investment needed for building ponds can be provided in the form of labor, which is not the case for cage, pens, tanks that require material to be purchased, or even imported. Supply of high quality feed used for exogenous feeding fish farming systems requires a high level of cash flow, which is not the case of semi-intensive fish farming systems using low cost inputs such as agricultural by-products and wastes from livestock and crops.

Expressed in man-days per unit of harvested fish, higher level of labor is required for fishpond system management (mainly for pond maintenance and cleaning, manuring, and harvesting) than for cages or other intensive fish production system. For a long time, it was admitted that endogenous feeding based fish farming systems were requiring only a low level of technical know-how from the fish farmers compared with exogenous feeding systems. The reality is far from so simple: the former system is developed in countries with an old tradition of fish culture practice and where a strong know-how has been accumulated even if mainly empirical.

The many attempts to transfer these fish culture models into countries where there was no fish culture tradition failed because on the other hand, high feed input aquaculture models are based on technologies that are easier to transfer because their main components (fish stocking density, feeding rate, and composition of feed,) are well defined and are carried out within a culture environment where the environment components either interfere only few (cage culture in lakes or rivers) or are under control (raceways, tanks). Level of risks, in terms of fish diseases, is considerably reduced in extensive culture systems than in intensive ones. Similarly, production costs and yields are higher in intensive systems than in extensive, and semi-intensive culture systems.

Ponds are multipurpose fish culture systems: they can be used for brood stock stocking and maturation, breeding according various methods (natural, semi-natural, artificial), fry nursing and fish growing out. Moreover, in pond structures such as hapas used for tilapia, fry production for example can be implemented. Intensive structures usually specialized for growing out purposes from fingerlings to market size fish.

3. The Main Cultivated Species in Freshwater Aquaculture

3.1 Carps

The main carp species cultivated in the world are primarily seven in number and are often grouped on the basis of their natural geographical occurrence: the so-called Chinese carps, which include the grass carp, Ctenopharyngodon idella, the silver carp, Hypophthalmichthys molitrix, and the bighead carp, Aristichthys nobilis, and the so-called Indian major carps, which include catla, Catla catla, rohu, Labeo rohita, and mrigal, Cirrhinus mrigala. The seventh species is the common carp, Cyprinus carpio. Taxonomically, carps belong to the family of Cyprinidae (order, Cypriniformes).

3.1.1 Grass Carp

Grass carp is a natural inhabitant of the Flatland Rivers of China, and the middle and lower reaches of river Amur in Russia. The fish has been introduced worldwide into over 50 countries since beginning of the twentieth century. In many countries, the main purpose of its introduction, in addition to culture, is biological aquatic weed control in natural waterways, lakes, man-made lakes and irrigation channels. Grass carp, like other Cyprinids, has a toothless mouth but has specialized pharyngeal teeth for rasping aquatic vegetation. The patterns of teeth structure of grass carp are adapted to the feeding habit. Large fishes are able to masticate the leaves of tough land plants such as fibrous grasses. Intestinal length relative to standard length from 1.6 to 2.7 in adults (0.5 in

7

larvae). Digestion of fiber in the grass carp is said to be incomplete and about half the food material ingested is excreted as feces, which, supposedly, can support directly or indirectly a large biomass of other species of fish in the fishpond.

The natural food of grass carp fry is first protozoa, rotifers and nauplii and then cladocera and copepods, with benthic algae when fry reaches 20-25 mm length. Then, the natural diet becomes phytoplankton and minute algae and, for fish above 30 mm, the natural diet becomes macro-vegetation. The preferred weeds according to the different regions of the world are: Wolffia, Lemna, Spirodela, Hydrilla, Najas, Ceratophyllum, Potamogeton, Vallisneria, and Myriophyllum. Under culture conditions, grass carps are often given substances like cereal brans, oilcakes, and silkworm pupae, and kitchen refuse as supplementary food.

In natural waters, grass carps reach a weight of 200 to 650 g at the end of the first year and, after 4 years the weight may be 4 to 5 kg. In the Yangtze River, fish weighing more than 20 kg have been caught. When cultured, growth is a function of many factors (stocking rate, feeding quality and quantity, competition with other fish when in polyculture conditions...). The range of daily growth of grass carp may be from 2.8 to 9.8 g.day -1 depending on culture conditions and climate environments.

Age at which grass carp attains maturity varies greatly according to climate and environment factors, mainly temperature: from 1–2 years in tropical Asia to 8–10 years in cold-temperature countries (Russia). Relative fecundity of grass carp varies from 80 000 to 120 000 ova per kg of female bodyweight (Table 4).

Table 4. Relative fecundity (number of ova per kg of female fish) in the main cultivated carp species.

The fish breeds during monsoon months in its natural habitat, the rivers, but does not spawn naturally in the stagnant waters of ponds and tanks. Induced spawning is required under culture conditions.

3.1.2 Silver Carp

Silver carp naturally occurs in the river systems Yangtze, West River, Kwangsi, and Kwangtung in South and Central China, and in the Amur Basin in Russia and the species has been introduced into many countries for aquaculture purpose.

The feeding habit of silver carp is mainly zooplankton, rotifers and nauplii of copepods at the young stages and expands as the fry grow to include copepods, cladocera and phytoplankton. Larger fry and adults feed on Flagellata, Dinoflagellata: primarily phytoplankton and secondarily zooplankton. The fish shows specific anatomical and morphological modifications in correlation with its dominant phytoplanktivorous feeding habit. The length of the gut of the adult fish is 15 times the body length. The gills of silver carp have a complex network and profusion of closely set gill rakers, allowing the fish to filter small size algal cells (down to 30 µm in diameter).

Under culture conditions, silver carp may reach 1.8 kg after 2 years, 4.6 kg after 3 years and usually, growth declines after the third year. The average growth rate is 6.3 g.day -1 during the 3 first years of culture.

As for grass carp, the age at first maturity of silver carp depends mainly on temperature. For example, in North China, the sexual maturity is attained after 5–6 years (weight: 2–5 kg) while in South China maturity is achieved only after 2–3 years (same individual body weight).

8

The average relative fecundity of Silver carp is slightly higher than for the grass carp. The fish breeds naturally during April-July in the flowing waters of its natural habitat, the rivers of China, but do not spontaneously in ponds and tanks where induced breeding is required.

3.1.3 Bighead Carp

The natural habitat of this carp is the same as for the silver carp and it has also been introduced into many countries. Larvae of bighead carp feed mainly on unicellular planktonic organisms, nauplii, and rotifers. Larger fry and adults go on feeding on various forms of plankton, mainly zooplankton. The alimentary canal of this fish is much shorter than that of silver carp, due to its feeding habit based on zooplankton, rather than on phytoplankton. The similarity in feeding habit between the bighead carp and the Indian major carp catla leads to a high degree of competition for food between these 2 species when grown together in polyculture in the same pond.

Under favorable culture conditions, growth rate of fingerlings can reach 6.3 g.day-1 and of adults as high as 14.7 g.day-1. Individual weight of bighead carp attains in these conditions 3.2 kg after 2 years culture and 10.7 kg after 3 years.

The age at first maturity of bighead is 3–4 years in South China (weight: 5–10 kg) and 6–7 years in North China (same weight). The fish spawns during monsoon in its natural habitat but does not breed naturally in captivity.

3.1.4 Common Carp

Depending on various authors, the common carp is native of the temperate regions of Asia, especially of China or would be rather native of the rivers draining into the Caspian Sea and the Black Sea or originated from central Asia and was introduced in ancient times into China and Japan in the oriental region and into Greece and Europe through Rome. The common carp has four subspecies and numerous varieties or strains. Among this amount of varieties, it can be found the “big belly carp” of the Kwantung and Kwangsi regions of China, the Indonesian orange-colored carp (Cyprinus carpio variety, Flavipinnis c.v.), and the mirror carp (Cyprinus carpio var. specularis) with its Aischgrunder (Germany), and Royale (France) varieties.

Common carp post-larvae start feeding on small zooplankton (Moina, rotifers, Cyclops, and nauplii). Ostracods, insects including chironomid larvae, Euglena, and Closterium are added when common carp fry reaches 20 to 100 mm long.

Common carp bigger than 10 cm feeds on decayed vegetable matter containing bottom dwelling organisms, mainly tubificids, mollusks, chironomids, ephemerids, and trichopterans. Common carp dig and burrow into pond dikes and bottoms, looking for organic mater: the fish takes in the mud from which digestible mater is sifted and the remaining rejected. This feeding habit results often in pond embankments deterioration and water turbidity.

The growth rate of common carp, depending on its culture conditions, is around 2 g.day-1 on average.

The age of common carp at first maturity ranges from 1 year (or even 0.5 year) in Tropical Asian countries (India, Indonesia, Thailand, and Malaysia) and in Israel, to 3–4 years in Europe. The average weight at maturity ranges respectively from several hundreds grams to 1.5-2.5 kg. Spawning season can occur year round (tropical countries) or only during a given period (May-June in Europe, March-August in Israel, and March-June in South USA) depending mainly on the

9

water temperature before and during breeding period. Common carp breeds naturally in its natural habitat, as well as in ponds and tanks. The eggs are adhesive and the fish requires adequate sub-strata for spawning purpose and attachment of eggs.

The relative fecundity of common carp varies from 90 000 to 300 000 ova.kg-1 of female bodyweight.

3.2 Catla

Catla is native of the freshwater sections of the rivers of northern India, Pakistan, Bangladesh and Burma. The species has been transferred into rivers of peninsular India and more recently to several countries of East and South East Asia.

Catla, as juvenile, feeds mainly on zooplankton (Crustacea) and as adult feeds for two thirds on zooplankton and for one third on algae. The feeding habit of the fish is mainly concerned with the water column but also with the bottom, attested by the occurrence of organic detritus, mixed with sand, mud and rooted aquatic plants in the gut.

Catla is the fastest growing of the Indian major carps. Under favorable culture conditions, catla can reach a weight of 3.2 to 4.1 kg after one year, 10.9 kg after 2 years, and 18 kg after 3 years.

Catla attains sexual maturity in ponds during the second year of life at an average weight of around 3.0 kg. The average relative fecundity ranges from 100 000 to 250 000 ova.kg-1 of females bodyweight when brood stock is over 3 years old.

The spawning season of catla occurs from May to August in North and Northeastern India, Bangladesh, and Pakistan. In South Indian rivers, the breeding season is more fluctuant (May-October) and may occur twice a year.

3.3 Rohu

The natural distribution of rohu is the same as for catla and the fish has also been transferred into peninsular India and many tropical countries all over the continents. The feeding habit of rohu is based on plant matter including decaying vegetation. The fish is a bottom and water column feeder and is much less adapted to take zooplankton than other Indian carps. Fry and juveniles of rohu, up to 100–200 mm long, feed mainly on unicellular and filamentous algae, and rotting vegetation, the percentage of the latter increasing in larger fish. Under culture conditions, rohu shows a very high growth potential though having a somewhat slower growth rate than catla. The best-recorded performances of rohu are reached 1.0 kg after one year and 2.6–5.4 kg after the second year of culture.

Rohu attains first sexual maturity at the end of the second year in ponds but it has been observed in India that the maturity can be performed in one year only (3–4 years in Bangladesh). The average weight at first maturity stands around 0.5 kg (length: 350 mm). The relative fecundity of rohu is reported to vary from 100 000 to 400 000 ova per kg of female bodyweight. The spawning season of rohu generally coincides with the monsoon, from June to September in most parts of the Indian subcontinent.

10

3.4 Mrigal

The natural distribution of mrigal is the same as for rohu and catla, and mrigal has been transplanted into waters of peninsular India, and of other Asian tropical countries for aquaculture purposes.

Mrigal feeds on detritus with a quite narrow range in food variety. It is a bottom feeder subsisting mainly on decayed vegetation but it can, occasionally, feed on the water column as an omnivore. Young fish feed mainly on plankton followed by semi-decayed organic matter and the reverse in the larger fish measuring over 560 mm lengths. Considerable amounts of sand and mud are observed in the gut of mrigal, representing up to 35% of the gut contents.

Under aquaculture conditions, catla can reach a weight of 1.8 kg after one year growing, 2.6 kg after 2 years and 4.0 kg after 3 years. Mrigal is reported in most cases to attain its first maturity when 2 years old in ponds (the males can attain their first maturity at the end of the first year). Relative fecundity of mrigal ranges from 100 000 to 400 000 ova.kg-1 of female bodyweight. The spawning season of mrigal corresponds to the Southwest monsoon in India, Bangladesh, and Pakistan and its duration varies in different regions of the Sub-continent from April to September.

3.5 Tilapias

The tilapias belong to the Tribe Tilapiini, an exclusively African group of fish within the family of Cichlidae. Previously regarded as members of a single genus, Tilapia, three main genera are now recognized, based on the last taxonomic revision (1983). In addition to anatomical characteristics, the criteria for generic distinction are based on the reproductive biology: Oreochromis (maternal mouth-brooders), Sarotherodon (paternal or bi-parental mouth-brooders), and Tilapia (sub-strata spawners).

Almost 100 species of fish are referred to by the common name tilapia but only three species feature significantly in aquaculture: Oreochromis niloticus, O. mossambicus, and O. aureus.

Nearly all the large tilapias most suitable for aquaculture belong to the Oreochromis group and within this group, Nile tilapia Oreochromis niloticus, is the most important species on account of its fast growth rate, adaptability to a wide range of culture conditions and high consumer acceptability. For these reasons, over the past 40 years, it has been transferred throughout the world to over 100 countries to become the mainstay of tilapia farming in many different culture systems, at all levels of intensification, from subsistence production to highly intensive farming.

All these tilapias species transfers, inside and outside Africa, as well as hybridization, which has been widely used in aquaculture lead to some confusion concerning the origin and the purity of the commercial strains. It is therefore most important to know the exact genetic make-up of the strains used for aquaculture production. Various biochemical and bio-molecular techniques (allozymes, mitochondrial DNA, and micro-satellites) have been developed for accurate genetic identification. Many of the problems associated with tilapia farming come from the exceptional mode of reproduction of the Oreochromis species. In addition to being a mouth-brooder fish, it attains its sexual maturity very precociously (within 6 months and even below 40 g of weight). This diversion of energy from growth into reproduction becomes a real constraint under culture conditions. Once mature, tilapia females produce multiple batches of eggs, the oral incubation of each egg batch being followed by only a short period of recovery before she is ready to breed again: each reproductive cycle lasts for about 1 to 1.5 month. On average, O. Niloticus produces from several hundred to about 2.000 eggs per batch and can breed up to 10 times in one year.

For aquaculture purposes, two main problems stem from this reproduction behavior:

11

• Spawning by Oreochromis tilapias is not synchronized among females, so that a population of brood fish will produce continuously, but at a low rate which is a key problem for the tilapia hatchery producers, combined with the cannibalism of young fry by the older recruits remaining in the ponds which is another major cause of declining fry output in brood stock ponds

• Precocious and continuous breeding of tilapias lead rapidly to overcrowding and stunting The main challenge for tilapia brood stock management is to develop conditioning and breeding systems as well as a very frequent removal of s wim up fry from breeding systems in order to increase the production of fry per unit area of hatchery pond.

Table 5. Bio-technical data on the production of Oreochromis niloticus fry in ponds according to various management practices in Asia and in Africa

To overcome the problems of tilapia overcrowding in ponds, several methods of controlling reproduction in tilapia culture systems have been developed: polyculture with a police fish eliminating the fry by predation and/or culture of monosex male fish (the males having a better growth rate than females). The methods adopted for all-male stocks production include hand sexing, hybridization and hormonal sex reversion. Hormonal sex reversal is now considered as the most efficient technique for all-male progenies production, at a commercial point of view, but is not without problems such as the level of technicity required to obtain 100% male populations and the possible impact of large-scale use of synthetic artificial steroid hormones to the aquatic environment.

In terms of feeding regime, Oreochromis tilapias, and particularly Oreochromis niloticus, are usually considered as phytoplanktivorous fish, able to ingest and digest large quantities of phytoplanktonic algae as well as blue green algae. According to several authors, tilapias would be the only real herbivorous fish considering their gut structure (length of 14 times the total length of fish and pH of stomach very acid, around 1.5 allowing the destruction of cells walls). Under culture conditions, in fertilized ponds, Oreochromis niloticus feeds on rotifers, copepods, Cladocera, chironomid larvae, diatoms, green algae, blue green algae, and decayed organic matter from macrophytes, nanoplankton as well as mud, clay, and sand. Oreochromis niloticus appears then to be an opportunistic fish in terms of feeding habit that is closer to omnivorous-detritivorous fish than strictly microphytophagous. The detritus fraction of feed ingested by Oreochromis niloticus appears to be most important and has been underestimated for a long time. The fish is able to adapt his morphology and his feeding behavior to digest the detritus component of feed: selection of organic fraction of feed and hydrolysis by using several small pharyngeal teeth, digestion thanks to the low pH of stomach and, at least, the amino acids are assimilated all along the gut.

Young Oreochromis niloticus tilapias feed mainly on micro-zooplankton and micro-crustaceans while larger fish feed on a wider range of food.

Oreochromis feed on the water column, on the bottom and on sub-strata following three main ways of ingestion: suction by creating a water flow directed into its mouth, filtration through its gill rakers, and grazing on sub-strata and on the bottom of the pond. Recent studies showed that grazing appears to be energetically the most efficient feeding strategy for Oreochromis tilapias.

The feeding habits of the main cultivated freshwater fish as well as fish species of potential interest for tropical aquaculture are given in Table 6.

Table 6. Trophic and spatial niches of freshwater fish species used in aquaculture.

12

4. Freshwater Fish Farming Management

4.1 Reproduction

Fish farmers need fry or “seed” predictably and in very large amounts. When the number of wild-caught fry falls below what farmers need to stock their ponds or cages, producers must induce adult fish to reproduce in captivity. Although it is possible to control fry production by using natural reproduction and environmental management, hormonal stimulated spawning is widely employed. With such a method, the better the reproductive physiology of fish is understood, the better are the chances of success.

The nervous and endocrine systems of fish act together to coordinate reproduction. Neural stimulation starts the chain of events, and latter links are hormonal. Stimuli from the environment (photoperiod, temperature, food etc.) are processed by sensory receptors; the resulting neural signal, when reaching the hypothalamus, induces the production of chemical messengers known as releasing hormones (gonadotropin-releasing hormone GnRH or luteinizing-hormone releasing hormone LHRH, which is a certain kind of GnRH). GnRH is small peptides (10-amino acid) and they are identical or only slightly different in most fish. They initiate the release of gonadotropin hormones by the pituitary gland (also named hypophysis). They are glycoproteins and they influence the production of sex steroids in the gonad itself. Sex steroids are responsible for gamete maturation and, if the appropriate environmental and social signals are present, ovulation (or spermiation) and spawning follow.

The gonadotropin release is inhibited by many biological systems, particularly hormonal ones. Although, there is some evidence for inhibitory action of a brain hormone, the neuropeptide Y (NPY), the strongest evidence for gonadotropin inhibition in fish is for dopamine.

Fish culturist wanting to induce breeding with hormones will first have to assess the maturity of breeders, as the success of the technology depend on accurate information about the state of the gonad. Fish readiness can be judged by considering the external appearance (large soft abdomen and a swollen gonad papilla for female, release of milt when abdomen is squeezed for males) or using more complex and time-consuming methods based on gonad biopsy and egg analysis (egg diameter and size distribution, morphology of oocyte, and localization of nuclei in cells).

Fish manipulations must be carried out in a way that minimizes the stress of reproducers: use of anesthetics, avoid of overcrowding, moist of hands and all cloth nets or holding slings before handling fish, cover of the fish eyes, minimization of noise etc.). Spawning hormone is generally administered intramuscularly or intraperitoneally, but some attempts were made to deliver them orally or by mixing them with a binding material (implantation). The last method permits a very slow release of hormone, over weeks or months.

The choice of hormone depends on many factors including species, cost and availability, egg incubation or larval-rearing facilities, and training. Fish farmers will use either a gonadotropin or a GnRH analogue, with or without a dopamine antagonist.

Hypophysation, the injection of crude fish pituitary extracts, has been used since the 1930s. It has many advantages and many drawbacks, as extracts are highly impure and contain accessory hormones and other components that may stimulate some fish and inhibit others. But, most experienced fish breeders agree that if good pituitaries are available, hypophysation is an excellent method of spawning. A typical modern hypophysation technique for freshwater fish involves two injections into females: a small dose stimulates germinal vesicle migration and is followed about 12 hours later by a larger one that induces germinal vesicle breakdown, ovulation and spawning. Males are generally injected only once, to induce sperm hydration coinciding with ovulation in the female. Some partially purified fish gonadotropins are also available, but their high cost limits their

13

use to research. HCG, the human chorionic gonadotropin can also be employed but the molecule is so different from fish gonadotropin that high dose must be used for many species and some may not respond at all. Moreover, the availability of gonadotropin hormones like HCG is low, the stability of hormonal products is weak and these compounds induce a prejudicial immunity response in several fish species.

On the contrary, many releasing hormones are available and they have three great advantages: they act early in the hormonal chain and cause the fish to produce its own gonadotropin, thereby eliminating all the problems caused by using a gonadotropin from another species. Second, the GnRH molecule itself is not highly species-specific. Third, they are simple, easily manufactured, stable molecules whose biological activity does not vary from lot to lot and, because they are active at such low concentrations, their use is economical. GnRH analogues can be used solely or together with a dopamine antagonist (pimozide or domperidone). Injecting a GnRH analogue followed by, or in combination with, a dopamine antagonist has been called the Linpe method, after Lin and Peter the researchers who started it. Effective doses of hormones vary widely but the trend is toward single injections of 5-20 µg.kg-1 GnRH analogue doses, although 1-100 µg.kg-1 have shown to be effective. Domperidone is usually effective at doses of 1-5 mg.kg-1. Table 7 shows a comparison of Linpe method and traditional hypophysation in induced reproduction of Indian carps.

Table 7. A comparison of induced reproduction results in female Indian carps, using carps hypophyses (CH) or Ovaprim® (commercial product containing 10 mg.l-1 domperidone and 20 µg.l-1 salmon GnRH analogue, by Syndel, Vancouver, Canada). * The differences between these two lines

come from unachieved ovulations (From Billard, 1995).

Fish that have been induced to ovulate and spermiate with hormones are often strip-spawned the gametes are removed by gently compressing the abdomen and then mixed manually. Artificial mixing of eggs and spermatozoa can produce very high fertility, over 90%, but techniques vary from fish to fish. The dry method is the best basis for fertilization. To make the most out of the very short period of sperm motility, mixing with eggs is done before any water is added. Quickly after, just enough of the natural spawning water or diluting solution is added to wet them thoroughly and after several minutes, fertilized eggs are washed with larger volumes of water, before being transferred to incubators (Zoug bottles).

Fish culturists have deadlines and production quotas, and persuading administrators and owners to put more effort into achieving natural reproduction, when hormone-induced spawning may be at least partially successful, will be difficult. But some of the best evidence that the natural method works, comes from the world of tropical aquarium fish where most of the breeding is done by natural means and successes are legion. In the case of the Asian arowana (Scleropages formosus), young fish are desired for display but can no longer be captured legally and, as a consequence, they command a very high price. In spite of this fact, the strategy that worked for reproducing it was simply to provide undisturbed space, water of the correct hardness and high quality diet. In fact, reproduction induced by management of environmental factors can be applied to many other species: carps (through a thermal stimulation, or by breeding fish in weedy ponds or ponds with artificial sub-strata/s), and osteoglossids (by breeding them in weedy pond, or by simulating flooding) etc. The development of new management methods for a better-controlled natural reproduction might be a very efficient tool for future.

4.2 Nutrition and Feeding

The performance and success of a fish production rely mainly on the quality of the feeding. In many technical models, in particular in raceways and cages where little, if any, natural food is available the fish farmer must provide a complete diet. These are usually diets rich in protein and vitamins, and they are therefore quite expensive. The costs of feed in intensive culture systems may reach 50% or more of total costs. Consequently, adequate nutritional practices have an ever-increasing

14

determinant role, not only in terms of economic optimization, but also in terms of maintenance of good health, improvement of reproduction or growth performance.

The protein nutritional requirements of fish are definitely higher than those of the mammals or the birds. The content of protein in food must be of 40–45% (dry weight) for salmon, 32–36% for the American catfish, 31-43% for carps, 44% for eel, 30–40% for tilapias, and using expensive fishmeal generally provides it. However, if it is true that fish have characteristics which explain why their metabolism depends more on the contribution of food proteins than the one of any other animal group (weak participation of the body proteins to the pool of amino acid precursors, importance of direct catabolism through the oxidative pathway of absorbed amino acids), the repercussion of these characteristics on the need in essential amino acids is hardly significant. On the whole, the latter is rather similar from one species to the other and is comparable with that of superior vertebrates, if one excluding the case of arginine, which can be synthesized by the cycle of urea. Under these conditions, the myth of the strict protein demanding fish must be abandoned. Moreover, even for such a long food chain fish as salmon, the percentage of retained protein from feed to edible food is as high as 30%, compared with 18 and 13% in chicken and swine respectively.

Tilapias are known to be much more herbivorous than catfish or carp. For that reason, several attempts were made to substitute fishmeal by plant protein in order to limit food cost. Replacement of fish meal by increasing proportions of algal meal showed decreased growth when more than 5% algae were included, as imbalances of amino acids or minerals weren’t compensated. However, under practical culture conditions in Israel, replacement of half the fishmeal by soybean has been found nutritionally equivalent, without any supplementation.

The lipid nutrition is one of the best-studied sectors in aquatic nutrition research. The main steps of the lipid metabolism are known. In spite of great similarities, they differ from those of superior vertebrates from various points of view. In particular, the aquatic environment is characterized by great amounts of polyunsaturated fatty acids in particular those with long chain (>20 atoms of carbon). Those of the series (n-3) are the ones for which fish have the highest needs, as opposed to the terrestrial superiors vertebrates. In the case of trout, the polyunsaturated fatty acids (n-3) must account for at least 10% of the food lipids and at least 14% in the case of carp. Moreover, the lipids can ensure a supply of energy. As the majority of fish badly digests the complex carbohydrates but catabolises the proteins for meeting their energy requirements, this characteristic makes it possible to save proteins and to reduce the cost of the feeding. Thus, the growth of trout can be improved without increasing the content of food protein by raising the lipid level. For this reason, lipids are nowadays employed more and more to manufacture very powerful feed. The food ration must contain lipids in significant but variable quantity according to the species considered. Salmon require 18–20% of lipid in their ration, whereas cyprinids need 7–18%. For tilapias, their requirements are lower than 10%. For most fish, the digestion is very high, since it is of 90%.

The third major component of fish nutrition is carbohydrates, which constitute a cheap source of energy. Generally speaking, fish do not have as large a capacity to use the carbohydrates as do the superiors vertebrates, although omnivorous and herbivorous fish like catfish, carps, and tilapia utilize it better than salmon. For the latter, carbohydrate excess can even entail an elevated mortality. The carbohydrates account for about 30% of the salmon ration and more than 50% of cyprinids. The digestibility of small carbohydrates (glucose, saccharine) is much higher (near 100%) than that of starch (70–80% generally, but sometimes < 50%), but the latter is the only carbohydrate that can be economically incorporated into feed. In fact, its digestibility increases with temperature, so that it is generally higher for tropical species than for temperate ones. Moreover, it is possible to improve it by cooking. Thus, the quantities of starch that are incorporated today largely exceed the limits, which were asserted two decades ago. Retained energy from feed to edible food in fish is 27% for salmon, compared with 12 and 16% for chicken and swine respectively.

Carbohydrates also contain more or less complex and weakly hydrolysable polysaccharides. The food role of these compounds is still badly known (stimulation of the digestive transit by some fibers

15

or delayed-action of stomach draining by others, exfoliation of the intestinal cells, increased elimination of cholesterol and minerals). Their food value is certainly negligible, even if the digestion of cellulose seems to be possible in the intestine of some herbivorous fish, due to the presence of bacteria. In tropical zone, many species have adaptations that enable them to better digest these compounds, in particular the detritus matter. Indeed, for the major part of fish, the organic detritus constitutes a temporary nutritive resource, when traditional food is not available. The slimming that often accompanies this qualitative change in feeding proves that these species are not physiologically able to exploit a detritus diet. But this general rule does not seem to apply in many tropical ecosystems where the species, in particular the Nile tilapia Oreochromis niloticus, are detritivorous.

The differences between fish and higher vertebrates are very tiny with regard to the requirements in vitamins and minerals. The A2 vitamin, another form of the retinol, is specific to them but it is quite as effective as vitamin A, cholin, and inositol are necessary, whereas calciferol and K vitamin have a limited role. Conversely, the E and C vitamins have a greater importance than that of superior vertebrates. Others food compounds can be added, in particular for the pigmentation of the flesh (salmon). These are the carotenoids. The effectiveness of their retention in flesh depends on many factors that vary according to the species and its age. If the pigmentation and transformation of carotenoids into vitamin A are well studied, the other nutritional aspects remain little known.

From this knowledge, it is possible to define formulas like those presented in Table 8. The calculation of the rations encounters three difficulties: the effective ingestion of distributed food, the pollution of the environment by unconsumed or undigested food and, the temporal variability of the fish nutritional needs, which requires a permanent adjustment of the distributed rations. The feeding ad libitum is difficult to estimate and is likely to cause losses. It can also be confronted with problems of energetic content. Thus, with low energy components, the ingested food can be limited by the volume of the stomach and in some cases, be insufficient to ensure a good growth performance. Conversely, with highly energetic food, the ingested ration can exceed the fish needs. The surplus is then stored into fat, which creates losses at the time of evisceration.

Table 8. Composition of food for freshwater fish in intensive breeding (G.kg-1, from Guillaume et al., 1999).

The rationing of animals must thus take into account parameters related to the fish (size, mass, and growth expected), to the food and to the environment (temperature). It is defined using the food conversion ratio (FCR, ratio between the quantity of distributed food and the weight gain of fish) and the growth rate. Food ration is calculated using former rearing results and indicative values of FCR given in rationing tables delivered with commercial feeds. The distribution can be carried out in a manual, automatic way or upon request. The manual distribution is by far the most frequent and most flexible method as it makes it possible to adjust the quantity of distributed food to the behavior of fish. It is however expensive in labor and constraining, even if the feeding is only practiced once a day. The automatic feeding allows the fractionation of the distribution, and even the uninterrupted distribution, making it possible to improve the food conversion ratio. Lastly, with the distribution upon request, fish come to take the desired quantity of food, using a suitable mechanical or electronic device (tactile stem). The disadvantage of this device is that it frequently exacerbates the competition among fish and thus increases the disparities of growth. Obviously, it is possible to combine the various modes of feeding, in order to benefit from the advantages of each one.

4.3 Fish Pond Ecosystem Management

4.3.1 Pond Dynamics

16

A fishpond is an ecological environment in which many bacterial, animal and plant organisms coexist and interact. Their impacts on the fish production are major, not only because of their food value but also because of the induced modifications of water parameters (oxygenation, pH). The contribution of natural feed to the fish yield is essential, even in systems based on the use of artificial feed. Table 9 shows the respective contributions to growth of feed pellets and natural food for fish, obtained using methods based on the measurement of the 12C–13C ratio in fish flesh and in its potential food components. Similarities in results allow the determination of consumed items (those with a 12C–13C ratio close to that of the fish flesh) and unconsumed ones (those with a 12C–13C ratio quite different from that of the fish flesh). Table 9 shows clearly that in such a pond, most of the growth comes from natural feed (> 50% for carp and > 60% for tilapia).

Table 9. Source of fish (3000 Cyprinus carpio.ha-1, 1000 silver carp.ha-1, 450 grass carp.ha-1 and 3500–7000 tilapia.ha-1) and prawn (5000–15000 Macrobrachium rosenbergii.ha-1) growth in

polyculture ponds as indicated by 12C–13C analyses in a pond receiving 25% protein content pellets at a daily rate of 2% of the tilapia biomass plus 6% of the common carp biomass reducing to 4% of the carp biomass when carps weight exceeded 150 g. (From Schroeder G.L. (1983). Sources of fish and

prawn growth in polyculture ponds as indicated by δC analysis. Aquaculture 35, pp. 29–42).

The fish production is carried out during cycles at the end of which the ponds are entirely drained and dried. As opposed to the perennial aquatic environments, the ecological colonization of ponds cannot be ignored because its consequences on the characteristics of the ecosystem are major. During this period, which lasts from a few days to a few weeks, many qualitative and quantitative changes contribute to structure the aquatic environment. The regulation processes are primarily endogenous factors, related to the dissemination capacities of the organisms. C/N ratio could strongly contribute to the aquatic settlement. Indeed, a medium rich in carbon and low in nitrogen might support the organotrophic behavior of bacteria instead of mineralization. Consequently it might promote the development of the bacteriophagous micro-zooplankton (protozoa, copepods nauplii, and rotifers etc). On the contrary, a nitrogen rich fertilization might support the mineralizing behavior of bacteria, and by the intermediary of phytoplankton, might stimulate macro-zooplankton. These steps, very unstable, precede a state of relative balance.

The ecological dynamics of balanced aquatic ecosystems are abundantly described in limnology (cascading interactions theory). In spite of that, there is not any model useful for ponds because these environments are too complex and variable to be correctly represented by simplified food chains. Any way, some information is available. All the energy transfers stemming from the primary production can be called autotrophic pathway. The planktonic micro-algae develop as a result of the photosynthetic activity. Thereafter, they are consumed by phytoplanktivorous fish (like the tilapia) or filtered by zooplankton, which is then consumed by fish. Actually the autotrophic pathway is much more complex, as the energy transfers are not linear at all. In Nile tilapia ponds, phytoplankton composition has a great importance. The biological filtration of small-size algae is not energetically interesting for tilapia, which then rather feeds on pond bottom when these algae dominate the aquatic flora. The nutritive value of these feeds is good, and fish growth is satisfying. Conversely, when large-size algae dominate (mainly cyanobacteria), filtration is energetically more interesting for fish, but the nutritive value of algae do not always meet tilapia's requirements. Moreover, water quality is frequently bad when cyanobacteria develop, and oxygen content is often very low. As a consequence, tilapia growth is not very high.

Other relations exist, that do not stem from the primary production. It is the heterotrophic pathway, identified using methods based on the measurement of the 12C–13C ratio in fish flesh and in its potential feeds (as above). Many works confirmed its significant contribution to the fish yield. The organic matter provides a source of carbon for heterotrophic grazing which then benefits the fish yield. Table 10 presents the relative contributions of the autotrophic and heterotrophic food chains to the growth of fish and freshwater shrimps in ground ponds receiving a high level of organic fertilization. If the growth of silver carp, a phytoplanktivorous species, relies exclusively on the autotrophic pathway, this is not the case of the omnivorous species like the tilapia, and especially the common carp whose growth relies for more than half on the heterotrophic pathway. However,

17

the heterotrophic pathway can also be unperceived when all the conditions are met to allow the autotrophic pathway to express its full potential (sunning, carbonates and minerals rich environment). Moreover, many autotrophic algae can have heterotrophic nutrition and the majority of planktonic organisms have a very diversified feeding. The biological mechanisms that contribute to the autotrophic and heterotrophic pathways are thus tangled and badly known.

Table 10. Relative contributions of the autotrophic and heterotrophic food chains to the growth of fish (3000 Cyprinus carpio.ha-1, 1000 silver carp.ha-1, 450 grass carp.ha-1 and 3500-7000 tilapia.ha-1) and freshwater shrimps (5000-15000 Macrobrachium rosenbergii.ha-1) in ground ponds receiving a high level of organic fertilization: 50-200 kg.ha-1.day-1. (From Schroeder G.L. (1983). Sources of fish and

prawn growth in polyculture ponds as indicated by δC analysis. Aquaculture 35, pp. 29–42).

In such a context of partial knowledge, the main objective of the fish farmers is to direct the circulation of energy to the pathways that best benefit the fish. The pond fertilization is the first tool at the disposal of the farmer. As a matter of fact, it is logical to assume that by stimulating primary production, one can stimulate the production of all other trophic levels and therefore affect fish yield. This can be done by using chemical fertilizers, which supply the minerals required for production of organic matter through photosynthesis. But when minerals are present in sufficient amounts, the density of phytoplankton increases to such a level that light or dissolved carbon soon become limiting factors. One way to overcome this limitation is stimulating the production of heterotrophic organisms by using organic fertilizers.

Figure 1 shows the relationship between growth rate (g.day-1) and the average weight (g) of common carp, as determined by two weeks interval samples weighing for four treatments. The comparison between curve 1 (no fertilization and no feeding) and curve 2 (fertilization but no feeding) shows the impact of fertilization, in prolonging the growth of fish to higher individual weights. As a matter of fact, when fish grow, the predation exerted on natural foods increases and sooner or later, the amount of natural organisms is insufficient to meet the fish nutritional requirements. The individual growth does not stop immediately, but its rate decreases quickly. In a fertilized pond, the growth evolution is identical, except that, as the natural feeds are more abundant, the critical level occurs at a higher individual fish weight.

Figure 1. The relationship between growth rate (g.day-1) and the average weight (g) of common carp, as determined by two week interval sample weighing for four treatments: (1) no fertilization and no feeding (black triangles); (2) fertilization but no feeding (empty triangles); (3) feeding on sorghum (black circle); and (4) feeding on protein-rich diet (empty circles). Each point is an average value

determined from four replicated ponds (from Hepher, 1988, redrawn).

4.3.2 Chemical Fertilization

Mineral aqua cultural fertilizers are classified into two categories: nitrogen and phosphorus fertilizers. In aquaculture, potassium fertilizers are not considered, as their fertilizing impact seems generally non-existent. Several nitrogen fertilizers are available on the market (urea, ammonium nitrate, ammonium sulfate, sodium nitrate, and calcium cyanamide). Those containing urea or ammonia have an acidifying effect on the environment that can be counter balanced by the use of lime. The fertilizing contributions are more effective when they are frequently applied at a low dose rather than when they are applied at a low frequency but high amount. The generally recommended amounts lie between 5–10 kg N.ha-1 every two weeks.

Various phosphorus fertilizers are available on the market. The spreading of a phosphorus fertilizer leads to phosphates release in water. The latter can precipitate when water is rich in calcium or if

18

the pH is high. Phosphates establish strong connections with the sediment, which makes them unavailable for the trophic network. For this reason, it is generally advised to dissolve fertilizers before spreading. The generally recommended amounts are between 8.75 to 17.5 kg P2O5.ha-1 every two weeks.

The use of mineral fertilizers is sometimes limited because of the toxicity of some compounds. Nitrogen toxicity is high when nitrites or non-ionized ammonia are abundant. The toxic NH3 component of ammonia becomes increasingly abundant when the pH rises.

Spreading unbalanced fertilizers can also promote the development of cyanobacteria, which are frequently very harmful to fish farming (decreasing of growth, water de-oxygenation that can involve fish mortality). As some of these algae can use the atmospheric N2, it is generally considered that they develop when the N/P ratio is weak, but proliferations were observed in an opposite situation. In fact, these algae seem to proliferate every time the environment is unbalanced (unbalanced N/P, ponds rich in iron etc).

4.3.3 Organic Fertilization

Just like in terrestrial agriculture, organic manures can also be used to fertilize the ponds. They generally consist of agricultural wastes and/or by-products, so that they are generally inexpensive. They effectively stimulate the ecosystem, as they do not only contribute to the nitrogen and phosphorus enrichment, but also to that of carbon and organic matter. That way, they do not only stimulate the autotrophic food chains, but also the heterotrophic ones. As a consequence, organic fertilizers allow reaching fish productions higher than those exclusively obtained with mineral manures. Table 11 gives the composition of some manure used in fish culture, but there may be considerable deviations from the values presented.

Table 11. Chemical composition of some organic manure. (from Delincé G. (1992) The Ecology Of The Fishpond Ecosystem, With Special Reference To Africa,

230 pp. Dordrecht, Netherlands : Kluwer Academic).

Trials were conducted to assess the respective effects of manures (layer chicken litter and dairy cow manure) and chemical fertilizers (urea + triple super-phosphate) on the production of Nile tilapia in earthen ponds. The results show that the better yields are obtained by using chicken manure (11.7 kg.ha-1.day-1) instead of chemical fertilizer or cow manure (8.6 kg.ha-1.day-1). The superiority of some organic fertilizers is confirmed by other experiments. In China, comparing chicken litter, pig manure and cow manure, the highest fish yields were obtained with pig and chicken manure (5.2 and 5.1 times the unfertilized control respectively), and the lowest with cow manure (3.9 times the unfertilized control).

On the other hand, their mineral content to dry matter ratio is 20 to 30 times weaker than that of inorganic fertilizers. Thus, 1 ton of organic manure is equivalent to approximately 45 kg of a 10:5:10 fertilizer.

A simple way to fertilize fishponds consists in breeding other animals over or near the pond. The problem is then to determine the livestock necessary to fertilize the pond. Table 12 shows some types of fish-animal associations and the productions that were obtained. The size of the associated breeding varies according to the level of intensification, to the weight and to the age of animals. For pigs, livestock generally recommended are 30 to 85 pigs.ha-1 and for ducks, 1000 to 3500 ducks.ha-1. The main constraint is related to the fact that the farmers have to master perfectly the two breeding.

19

Table 12. Some fish-animal associations and related fish production (t.ha-1.year-1).

The dissolved oxygen concentration results from a balance between the phytoplankton autotrophic activity (main source of oxygen) and the bacterial heterotrophic one (main oxygen consumer). There is a typical daily evolution, characterized by a progressive oxygenation during the day (oxygen production by photosynthesis > consumption by bacterial activity) and a de-oxygenation during the night (oxygen consumption only, no photosynthesis). As already seen, organic fertilizations stimulate the heterotrophic, oxygen consuming, activity. When in excess, they can involve total de-oxygenation and mortality of all aquatic organisms. For that reason, the maximum recommended spreading quantities are 100 to 150 kg of dry matter per hectare and day. More precisely, the recommended daily contributions (expressed as a percentage of fish biomass) are 3 to 4% for the bovine and liquid pig manure, 2 to 4% for the chicken litter, and 2 to 3% for the duck litter.

Some ponds do not answer correctly to fertilization because of their low pH. Indeed, acid mud strongly absorbs phosphate. Moreover benthic organisms, in particular bacteria, do not develop correctly when pH is too low and phytoplankton lacks carbon and calcium in low hardness and low alkaline water. To increase the benefit of fertilization, pH of mud must be comprised between 6 and 7 and total alkalinity must be at least 20 mg.l-1 of CaCO3. Lime spreading makes it possible to solve these problems, and they have other advantages. Thus, they make it possible to sterilize the pond during drying by eliminating the parasites and pathogens; they improve the decomposition of the organic matter; they accelerate the nitrification; they reduce the sediment redox potential; they allow the suspended or dissolved organic matter to flocculate and sediment and thus improve the light penetration in water. Lime can also be brought to compensate the acidifying effect of some nitrogen fertilizers. In this case, the necessary amounts are proportional to the quantity of fertilizer spread. They are generally comprised between 85 and 170 kg for 100 kg of acidifying fertilizer.

Lime must be spread on the bottom in drained ponds or at the water surface in full ponds. Agricultural lime spread quantities are generally comprised between 2000 and 10000 kg.ha-1 and the operation must be repeated every 3 to 5 years. It is recommended to bring each year a half of the initial contribution.

4.4 Fish Populations Management

4.4.1 Polyculture and Monoculture The fertilization is not very effective if stocking is not well carried out. Fish population control must be total, which implies to limit the introduction of wild species (drain at the water entrance) and to control the reproduction inside the pond (monosex breeding, hormonal treatment, or polyculture with a police fish).

Polyculture consists in associating fish species w ith complementary diets, in order to increase the biomass produced by the pond, and if possible, to benefit from synergic affects between the different fish species. It is an intermediate situation between monoculture where the radiant energy flux is concentrated on one species and the balanced natural ecosystem where the recipients of flow are very numerous. Table 13 proposes a classification of the various types of fish farming polyculture practiced around the world.

Table 13. Typology of fish farming polyculture practiced around the world.

20

If polyculture originates from Asia, it is now practiced on all continents. Chinese use two methods. The first one consists in realizing the productive cycle in a series of ponds stocked with different sized fish. In the other, the breeding is carried out in the same pond, until the commercial size is reached. The fish are harvested as soon as they reach the market size and are replaced by fingerlings, at least for a period of the year. The species and ratios used are presented in Table 14.

Table 14. Stocking density of some polyculture ponds, as a percentage of the different fish species.

In traditional polyculture of Eastern Europe, carps remain the dominant species and the other fish hardly represent more than 10%: tench (6–7%), pike (0.9–1.7%), salmon (3.2–5.5%), others (1.2–4.8%). Because of their growth performances and their short food chain feeding regime, the herbivorous Chinese carps (Ctenopharyngodon idella, macrophytophagous, and Hypophthalmichthys molitrix, phytoplanktivorous) are frequently included in the polyculture. Israeli polyculture associates 3000 common carps (Cyprinus carpio) with 1000 silver carps (Hypophthalmichthys molitrix), 500 grass carp (Ctenopharyngodon idella) and 7000 tilapias. In Africa, the mixed-fish farming associates the Nile tilapia (Oreochromis niloticus) with catfish (Heterobranchus isopterus, Clarias sp.), one osteoglossid (Heterotis niloticus) and the predator Hemichromis fasciatus (to eliminate undesirable fry), according to the ratio 0.03 Heterotis niloticus, 0.04 Heterobranchus isopterus, and 0.2 Hemichromis fasciatus for each tilapia. Under these conditions, the secondary species can increase the total fish yield by more than 40%. In South America, experiments were led with Colossoma macropomum as main species and Prochilodus sp., Cyprinus carpio and tilapias as secondary ones. However, the South American references describing powerful associations are rare and the practices used are rather the result of the empirical experiment of the fish farmers.

Benefits of polyculture are diverse:

• Better and complete utilization of natural feed, as a fish species, even with a wide food spectrum, does not fully utilize all pond trophic resources

• Avoidance of some trophic deadlocks. When a dense stock of common carp is raised in monoculture, a small crustacean, Bosmina longirostris develops and it is considered a harmful side effect as this crustacean feeds on phytoplankton and is not grazed by common carp. In this way, Bosmina longirostris stands as a competitor for other herbivorous zooplanktonic organisms, which otherwise would be consumed by common carp. But when silver carp is introduced in polyculture with carp, Bosmina longirostris declines as a consequence of the grazing of Hypophthalmichthys molitrix

• Enhancement of natural food. The common carp stirs the pond bottom for feeding purpose and this behavior resuspends and aerates the sediment, oxidizes organic matter and improves the recycling of nutrients that stimulates the production of natural food

• Double fertilization. Dejections of herbivorous fish (H. molitrix, C. idella) are so rich that their fertilizing impact can be compared with that of an associated terrestrial breeding. This effect is sometimes called the “double fertilization” because a chemical fertilization is much more effective when these fish are in the polyculture. This “double fertilization” can improve the carp yield by 14–35% compared with the “normal fertilization” observed in monoculture ponds

• Improvement of water quality. In pond, improvement of oxygenation occurs due to the presence of silver carp or tilapia. Silver carp consume excess algae which otherwise could create an imbalance between production and consumption of oxygen. Tilapia may also improve oxygenation by consuming bottom organic matter that would otherwise have been mineralized by oxygen consuming bacteria

• Control of undesirable organisms. Mollusks control is possible in fishponds by using 75–100 black carps.ha-1 or 200 Heterotis niloticus.ha-1, whereas small wild fish or shrimps proliferation’s can be eliminated by using 200–600 carnivorous fish.ha-1

On the other side, there are some negative effects. They usually consist in competition among the different species when an imbalance is created. For example, polyculture of Colossoma

21

macropomum with Piaractus brachypomus and/or Brycon sp. lead to poor growth rate, probably because of competition for the same food by the various species.

But, when stocking density is very high, the relative role of pond productivity in the overall nutrition of fish decreases since existing natural food resources have to be divided among more individuals. The gain that can be obtained by polyculture is relatively small, while the extra work involved with sorting the different species of fish at harvest time becomes a burden. Monoculture is therefore the only method of culture used in running water systems and in cages where the supply of natural food is limited. In ponds, high stocking densities are not common, as oxygenation is usually a limiting factor. But as it is possible to aerate pond water, some methods have been developed. In a recent paper, a monoculture of Oreochromis niloticus proved to have a production potential higher than a 4:3:3 polyculture of Hypophthalmichthys molitrix, Labeo rohita, and Cirrhinus mrigala at 1 and 3 fish.m-2. The main explanation for this result was that dissolved oxygen concentration in the monoculture tilapia pond might have allowed an increase of the fertilization, whereas this parameter was critical in the carp polyculture.

4.5 Pond Stocking

The optimum stocking density of a fishpond is that amount of fish released into the pond at the beginning of the production period, which guarantees the highest possible economic income. The assessment of the fishpond stocking is one of the most important parameters for making the success of the breeding.

For aquacultural systems, a crop of young fish will likely grow at a near maximum rate until food or other environmental conditions become limiting. This point is termed the critical standing crop (CSC). Even though growth is reduced at CSC, biomass continues to increase once fish exceed CSC, until the population reaches the carrying capacity (K). At K, density effects of the population are so strong that growth reaches zero and biomass remains stable. As long as the rate of increase in fish density is higher than the rate of decrease in individual growth rate, yield increases. But when the decrease in growth rate exceeds the increase in fish density, yield decreases, as shown in Figure 2.

Figure 2. Schematic presentation of the relationships between the stocking density, the short interval growth rate and the short interval yield per unit area, with (broken line) and without (solid line)

supplementary feeding (from Hepher, 1988, redrawn).

For example, in a carp pond fertilized and fed with cereal grain (2200 carp.ha-1), CSC will be reached when the fish attain an average weight of 275 g (0.275 x 2000 = 550 kg.ha-1). Below this weight the potential growth rate has been achieved, but above 275 g their growth rate will be less than the possible potential due to lack of food. When the fish attain 1 kg (2000 kg.ha-1) they will cease growing. If however the same pond is stocked with 4000 carp.ha-1, they will reach CSC at 137 g and will cease growing at 500 g.

If fish are stocked in ponds at low density and natural food are abundant, they will grow at a maximum rate for that temperature. Addition of supplemental feed at this point has no effect, because food is not limiting. However, once stocking reaches CSC, food becomes limiting. Growth then diminishes unless management is intensified. If natural food production can be enhanced by fertilization, growth should again increase up to a new CSC, at a higher level. At that point, supplemental feed may be required to increase growth to maximum. Again, another CSC will be reached until food quality or water quality limits fish growth. This relationship between fish stocking and input level can be clearly seen in figures 1 and 2.

22

The density can be used for regulating the average growth rate of the fish and therefore, the length of the rearing period. As formerly seen, when stocking rate is increased, CSC will be reached at a lower individual fish weight and the growth rate above CSC will be reduced. The average growth for the entire rearing period will thus be lower. More widely, yield and individual growth rate are respectively positively and negatively correlated to density. In other words, up to a certain level, the lower the density, the quickest the growth and the lowest the yield.

The economic income of a fish farm is not only dependent on the total yield, but also on the selling price of the fish. When the fish is sold at the same price whatever its individual weight, the fish farmer will choose the optimal density in which the fish utilizes the natural food to give the highest yield per unit area. Larger fish are sold at a better price per kilogram on the market; the fish farmer will have to find a compromise between yield and final individual weight. In semi-intensive culture conditions (use of manure or low value agricultural by-products), the main factor under his control is the stocking density. By using low densities, he may have better growth rate, higher final individual weight but a lower total fish yield. With a higher growth rate, the duration of growing cycles is shortened, which may give a better cash flow. Experiments carried out in Côte d'Ivoire (Figure 3) showed that by using rice bran as unique input, the compromise between yield and average final weight is situated at a density comprised between 4000 and 7000 Nile tilapias.ha-1. There is now a strong tendency for low input aquaculture development to advise African fish farmers to use lower stocking densities than practiced before (20 000 fish.ha-1).

Figure 3. Yield and average final individual weight of Oreochromis niloticus as a function of stocking density.

5. Freshwater Fish Farming Economics

As it has been seen, all levels of intensification in freshwater aquaculture are developed around the world: from the most extensive to the most intensive. Two examples are presented here.

In the Philippines, the annual budgets of several fish farming systems are given: rice-fish culture, fish-pig integrated farming, and floating cage fish culture (Tables 15–17).

Table 15. Simple annual cost and return for a 1.000 square meter rice-fish culture project (two croppings) in the Philippines as of 1990 (in P, 1 US$ # 30 P).

Table 16. Semi-annual budget for a 1.000 square meter excavated earthen fishpond integrated with a head pig fattening, as of 1990 (in P, 1 US$ # 30 P).

Table 17. Annual budget for a tilapia cage culture farm in the Philippines.

Economic analysis shows higher profitability for integrated systems. A comparative analysis of rates of return of various systems shows that integrated livestock-fish systems are the most profitable with rate of return on investment higher than 63%.

Table 18. Rates of returns of different types of farming in the Philippines 1990.

23

Integrating fish with rice increases income by about 30–40% the same increase is obtained by integrating fish with livestock. In the latter, however, the contribution of fish to net income may be higher than 50%.

Cage culture, the most intensive farming system, needs a high level of cash flow and operation costs consist mainly of feed and secondarily of fry. This fish culture system leads to the lowest net profit per kg of fish produced. Hence it requires larger production units for a given global net income for the fish farmer.

In West Africa, tilapia cage culture in Niger shows the same economic trends as analyzed for the same culture system in the Philippines in terms of production costs.

Table 19. Financial analysis for a 20-m3-tilapia culture-floating cage in Niger (West Africa) as of 1995 (in FCFA, 1 US$ # 600 FCFA).

In Cote d’Ivoire a comparative analysis of various pond fish farming systems puts into evidence that the most intensive level of pond management, using a balance diet, does not lead to the highest financial efficiency.

Table 20. Economic efficiency ratios for different fish culture models developed in Cote d’Ivoire (West Africa) (in FCFA, 1 US$ # 600 FCFA).

This analysis shows also the advantage of using a combination of manure with supplementary low cost feeding (rice bran) in the case of M3 model.

If fish culture has to be compared to other agricultural activities in terms of economic efficiency, the crops that are competing with fish culture for land and water use are essentially irrigated rice, maize and vegetable. Comparison of the monetary products indicates that performing fish culture models such as developed in Cote d’Ivoire, make a more efficient use of land than irrigated rice: 13.600– 22.000 FCFA.100m-2.year-1 for fish culture against 1600–5600 FCFA. 100m-2.year-1 for irrigated rice.

In terms of the efficient use of family labor, monosex tilapia culture models compare favorably with food crops: 2500–3900 FCFA.day-1 for fish culture, 460-2000 FCFA.day-1 for irrigated rice and 670–1100 FCFA.day-1 for yam culture. In addition, fish farming gives higher economic efficiency rates than generally obtained for other categories of agricultural projects (10–20%). Moreover, the rates of return of capital of efficient pond fish farming observed in Africa, ranging from 43 to 52% are totally comparable with those recorded in the Philippines (48–63%).

6. Conclusion and Prospects

Although the total production of finfish and shellfish from capture fisheries amounted to 92 million mt in 1995, only 61 million mt (live weight) or 66% was available for direct human consumption as “food fish”. The remainder (31 million mt) was reduced into fishmeal and fish oil for use in animal feeding or for industrial purposes. Between 1984 and 1995, the volume of capture fisheries grew at an annual rate of 1.5%.

On another hand, aquaculture has been the world's fastest growing food production system with food fish production increasing at an annual rate of 10.9% over the same period, compared with 3.1% for terrestrial livestock meat production. Aquaculture’s contribution to total world food fish

24

landings has increased more than two fold between 1984 (11.5%) and 1995 (25.6%). Contribution of aquaculture to world total finfish production in 1995 reached 23% (70% of total freshwater finfish, 37% of total diadromous finfish and only 1.3% of total marine finfish). In terms of food supply, aquaculture produced 6.2% of the total world farmed animal meat production, ranking fourth in terms of global meat production (pig: 37.6%, beef and veal: 24%, and chicken meat: 20.9%).

Over 85% of total aquaculture food fish production came from developing countries and particularly from low-income food deficit countries, which supplied over 76% of total food fish output from aquaculture. In these countries, per capita aquaculture fish production increased from 1.2 to 4.5 kg between 1984 and 1995. The contribution of aquaculture (and particularly freshwater finfish aquaculture) to rural food security in developing countries is most probably much greater than reported in official country statistics because self-consumption by fish farmers and their families is not recorded.

In most leading aquaculture-producing countries in Asia, food fish plays a major role in human protein nutrition by supplying more than one third in the total animal protein intake: 35% in Vietnam, 41.1% in Thailand, 51.5% in the Philippines and up to 65.2% in Koran DPR. In Africa, although the continent had the lowest contribution to world's aquaculture production (below 0.4%), food fish play an essential role in supplying over 30% of the total animal protein intake of populations in countries such as Ghana (58.6%), Congo (45.3%), Malawi (44.2%), Senegal (37.8%), and Cote d’Ivoire (36.0%) among many others. This strong tradition of fish consumption should help aquaculture to develop in the forthcoming years. In Latin America, a specific characteristic of aquaculture is that it is mainly export oriented (shrimp and salmon) and freshwater fish culture accounts for only less than 20% by volume of total aquaculture production in 1995. By contrast, the bull of aquaculture food fish production in developed countries is generally restricted to the production of higher value food fish species for luxury or export markets, and concerns mainly marine or diadromons fish and shrimp.

For the future, it is expected that freshwater aquaculture will continue to provide the major part of output. The main species will most probably be the lower-value herbivorous and omnivorous finfish (and shellfish) that feed low in the aquatic food chain. These species are grown mostly in ecologically efficient and environmentally benign polyculture systems, less demanding in terms of inputs, widely integrated into the agriculture production system. This form of aquaculture will continue to supply substantial quantities of fish protein for large segments of the population in many developing countries and will be, most probably, the primary area of development for lower cost production.

In terms of fish biodiversity used for aquaculture purposes, the number of cultured fish species (taxa), increased by 34% between 1984 and 1994. However, freshwater fish culture production remains largely dominated by only 9 species, which account for 78% of the total production. All these species are herbivorous or omnivorous and feed low in the food chain. For example, in Latin America, in countries belonging to the Amazon River basin, in spite of the enormous existing genetic resources potential, very few native freshwater fish are being culture. Two trends seem to emerge at the beginning of third millennium: the look for fish species diversification requested by most fish farmers around the world and the need to improve the majority of farm-raised aquatic animals that are still very similar to their wild forms. Differently expressed, what will prevail in the future: will emphasis be put on genetic improvement of already cultivated species for which culture technologies are mastered or on domestication of “new” species of aquaculture interest among the 25 000 existing fish species in the wild? Or both? The limited existing research potential will need to choose for priorities, knowing that both processes have their controversial issues. Genetic improvement by means of gene manipulation and gene transfer is perceived as having a high level of risk for wild species and global environment as well as for consumers. The emergence of newly domesticated species may lead to increasing introductions outside of their natural range into countries willing to diversify their fish production.

In terms of fish feeding, it appears that most commercially available aqua feeds for extensive and semi-intensive pond farming systems are over-formulated irrespective of the potential natural food availability. Therefore, tremendous research work is still required to understand the mechanisms

25

prevailing in fish culture ponds in terms of food web leading to fish production, fish interactions within polyculture and stocking rates with regard to optimal inputs aiming at minimizing feed wastage then reducing production costs, and maximizing benefits.

In terms of sustainable development, emphasis should be laid in the future on farming systems than can contribute positively to environmental improvement. Recycling of nutrients and organic matter as well as many types of wastes through integrated farming systems are long recognized as being environmentally sound. Rice-fish culture can help farmers reduce use of environmentally dangerous pesticides; waste-water-fed freshwater aquaculture can be used to recover excess nutrients from livestock sewage, night soil or agricultural by-products and wastes. Freshwater aquaculture thereby contributes reducing risks of eutrophication and pollution. Negative effects of aquaculture on the environment have been mainly associated with high-input, high-output intensive systems (e.g. cultures of salmon in raceways and cages).

Bibliography

Billard R. (1995). Les Carpes: Biologie et Elevage, 387 pp. Paris, France: INRA. [In French, this book gives a very wide description of carps biology and breeding. English readers may also consult Aquaculture Of Cyprinids by the same author, same editor.] Egna H. S. and Boyd C. E. (1997). Dynamics Of Pond Aquaculture, 437 pp. Boca Raton, USA: CRC Press. [A complete review of pond ecology and management of trophic food webs.] FAO (1997). Review Of The State Of World Aquaculture. FAO Fisheries Circular. No.886, Rev. 1. Rome, Italy: FAO Inland water resources and aquaculture service, Fishery resources division. [An update of the regular FAO reviews of world aquaculture.] Guerrero III R. D. (1987). Tilapia Farming In The Philippines, 85 pp. Manila, Philippines: National Book Store. Guillaume J., Kaushik S., Bergot P., and Metailler R. (1999). Nutrition et Alimentation des Poissons et Crustacés, 489 pp. Paris, France: INRA/IFREMER. [In French, an up-to-date complete description of fish nutrition.] Harvey B. and Carolsfeld J. (1993). Induced Breeding In Tropical Fish Culture, 144 pp. Ottawa, Canada: IRDC. [A useful handbook for induced fish reproduction.] Hepher B. (1988). Nutrition Of Pond Fishes, 388 pp. New York, USA: Cambridge University Press. [The reference in nutrition of pond fishes.] Hepher B. and Pruginin Y. (1990). Commercial Fish Farming, 261 pp. New York, USA: John Wiley and sons. [This presents a wide description of technologies used in commercial fish farming.] Jinghran V. G. and Pullin R. S. V. (1985). A Hatchery Manual For The Common, Chinese, and Indian Major Carps, 191 pp. Manila, Philippines: ADB/ICLARM. Lazard J., Morissens P., Parrel P., Aglinglo C., Ali I., and Roche P. (1990). Méthodes Artisanales D’Aquaculture du Tilapia en Afrique, 82 pp. Nogent sur Marne, France: CTFT-CIRAD. [A complete review of the technologies used for tilapia production in Africa.] PCAMRD (1990). Integrated Crop-Livestock-Fish Farming Systems. Los Baños, Laguna, Philippines: PCAMRD. Pullin R. S. V., Lazard J., Legendre M., Amon Kothias J.B., and Pauly D. (1996). Le Troisième Symposium International Sur Le Tilapia en Aquaculture, 630 pp. Manila, Philippines: ICLARM/CIRAD-EMVT/ORSTOM/CRO. [Proceedings of the international symposium on tilapia in aquaculture. Also available are proceedings of ISTA II and ISTA IV.]

Biographical Sketches

Jerome Lazard graduated from Montpellier University (Ph.D. Aquatic Ecology) and Paris 12 (HDR). He is currently appointed to the French government-owned Centre de Coopération Internationale en Recherche Agronomique pour le Développement (CIRAD) in Montpellier (France) where he is in charge of the Aquaculture Research Unit. He spent 15 years as a scientist in West Africa where he implemented and carried out R. and D. fish culture projects, mainly based on tilapia aquaculture. Once back in France, he did several missions both as a consultant for building research and development strategies in the field of inland fish culture in many tropical countries (Africa, Asia, South America) and for the implementation of scientific collaborative programs in this field. His research topics focus on tropical fish culture production systems. He has been awarded the silver-gilt medal of the French Academy of Agriculture and he his a Life-Member of the Society of Aquaculture Engineers of the Philippines. He is the author of publications on tilapia aquaculture and fish culture projects analysis.

26

Lionel Dabbadie, post-graduated from the Agronomic School of Montpellier (France) with a doctorate from the University of Paris 6, has been working on fish farming in Africa and Brazil for the French government-owned Centre de Coopération Internationale en Recherche Agronomique pour le Développement (CIRAD). He has been awarded the silver medal of the French Academy of Agriculture for his work on the pond dynamics in the framework of African extensive fish farms. He is presently working in the Brazilian Tocantins State on a project of culture of Amazonian native fish species, particularly on the endangered Osteoglossid Arapaima gigas, in cooperation with the Tocantins government and private operators.

1

Appendix 1: Scientific and Common Names of Species Cited

Aristichthys nobilis Big head carp Bosmina longirostris Bosmina Brycon sp. - Catla catla Catla Ceratophyllum Hornwort Cirrhinus mrigala Mrigal Colossoma macropomum Tambaqui Ctenopharyngodon idella Grass carp Cyprinus carpio Common carp Cyprinus carpio var Flavipinnis c.v. Indonesian orange colored carp Cyprinus carpio var Specularis Mirror carp Dinoflagellata sp. Dinoflagellates Euglena sp. Euglena Flagellata sp. Flagellates Hemichromis fasciatus Banded jewelfish Heterobranchus isopterus Azu isii Heterotis niloticus Heterotis Hydrilla sp. Hydrilla Hypophthalmichthys molitrix Silver carp Labeo rohita Rohu Lemna sp. Common Duckweed Macrobrachium rosenbergii Giant freshwater prawn Moina sp. Moina Myriophyllum sp. Water milfoil Najas sp. Water-nymph, Spiny Naiad Oreochromis aureus Blue tilapia Oreochromis mossambicus Mozambique tilapia Oreochromis niloticus Nile tilapia Piaractus brachypomus Pirapatinga Potamogeton sp. Sago Pondweed Prochilodus sp. Curimbata Scleropages formosus Asian bony tongue Spirodela sp. Duckweed Vallisneria sp. American Wild celery Wolffia sp. Duckweed

1

Table 1. Different levels of intensification of fish farming systems.

Fish stocking density (pieces.m-2) < 1 1–5 5–10 10–100

Farming system infrastructure Pond Pond Pond Pond/cage Raceway/pond/pen

Yield (t.ha-1.year-1) 0 - 1 1 - 5 5 - 15 15 - 50 > 50 and up to 100 kg.m-3

Management No input Low quality manure macrophytes

High quality manure, pellets

Pellets, aeration, and recirculation

Pellets, high level of water recirculation (natural or artificial)

Degree of intensification Extensive Semi-intensive Intensive Super intensive

1

Table 2. Stocking rates for Nile tilapia in cages of different sizes and management schemes in natural productive lakes in the Philipines.

Cage size* (m²) Number of fish.m-2 Number of fish per cage Management

1 (1×1) 200 20 With feeding

25 (5×5) 100 2500 With feeding

100 (10×10) 50 5000 With or without feeding

400 (20×20) 25 10 000 With or without feeding * Cage depth of 2 meters in average

1

Table 3.Types of fish culture systems according to the different feeding management practices

Production factor Endogenous feeding (Mainly ponds)

Exogenous feeding (Mainly cages, pens,

and raceways) Land - + Water Water area Water flow Impact of aquaculture on environment + - Capital/cash flow + - Labour (expressed per kg of fish produced) ± ± High quality feed + - Technicity level of fish farmers - - Level of risk + - Production costs + - Yield per surface (or volume) unit - + Multi-purpose of culture structure + -

+: Advantage for development -: Constraint for development

1

Table 4. Relative fecundity (number of ova per kg of female fish) of the main cultivated carp species.

Carp species Number of ova.kg-1 of females

Grass carp 80 000–120 000

Silver carp 160 000–195 000

Bighead carp 70 000–130 000

Common carp 90 000–300 000

Catla 100 000–250 000

Rohu 100 000–400 000

Mrigal 90 000–420 000

1

Table 5. Bio-technical data on the production of Oreochromis niloticus fry in ponds according to various management practices in Asia and in Africa.

Pond surface area (m²) 4 500 350 200 100

Broodstock density (ind.m-²) 0.16 0.7 4 0.6 Mean weight of broodstock (g) 62–356 100 (&)/240 (%) 80–100 100–150

Sex ratio (& : % ) 3: 1 3: 1 3: 1 3: 1 Duration of culture (days) 250 120 45–60 62 First harvest of fry (days after stocking) 60 35 14 12

Harvest intervals (days) 30* 15*

6 times/day at 2-hour intervals starting at 0700**

4 to 6 times/day when water is warmest

Feed/Fertilization Organic and

mineral fertilization

50% RB+ 50% PC

Organic fertilization + RB (75 %) + FM (25 %)

Rice bran (1.5 % bodyweight/day)

No. Of fry produced (ind.m-2.month-1) 8.0 45.4 200–250 400 Mean weight of fry produced 4.3 g 0.7 g A few mg to 0.1 g A few mg

Country Philippines Cote d’Ivoire Philippines Niger * Using seine nets. * * Using small-mesh hand nets RB: Rice bran PC: Peanut cake FM: Fish meal

1

Table 6. Trophic and spatial niches of freshwater fish species used in aquaculture.

Trophic niche Spatial niche Filter feeder Macro-particulate Scientific name Family Common name

Phytoplankton Zooplankton Macrophyte Detritus/

Invertebrates Predacious Surface Bottom Column

Ctenopharyngodon idella Cyprinidae Grass Carp ** + Hypophthalmichthys molitrix " Silver Carp ** * +

Aristichthys nobilis " Big Head Carp * ** +

Cyprinus carpio " Common carp * ** +

Labeo rohita " Rohu ** * +

Catla catla " Catla * ** +

Cirrhinus mrigala " Mrigal ** +

Oreochromis niloticus Cichlidae Nile tilapia ** * +

Colossoma macropomum Serassalmidae Tambaqui * ** +

Heterotis niloticus Osteoglossidae " ** * * +

Clarias gariepinus Clariidae African Catfish ** * +

Hemichromis fasciatus Cichlidae " **

Cichla oscellaris " " **

Channa striata Channidae Snakehead ** ** Major trophic niche * Minor trophic niche

1

Table 7. A comparison of induced reproduction results in female Indian carps, using carps hypophyses (CH) or Ovaprim® (commercial product containing 10 mg.l-1 domperidone and 20 µg.l-1 salmon GnRH analogue, by Syndel, Vancouver, Canada). * The differences between these two lines

come from unachieved ovulations (From Billard, 1995).

Catla catla Labeo rohita Treatment Ovaprim® CH Ovaprim® CH Temperature (°C) 25–31 25–31 27–30 27–30 Dose (kg-1) 0.4–0.6 ml 20–24 mg 0.25–0.35 ml 18 mg Number of females 74 68 68 67 % Total ovulation* 93 78 100 90

% No response* 0 12 0 0 % Fertilization 83 77 95 83 % Hatching 95 92 95 90

1

Table 8. Composition of food for freshwater fish in intensive breeding (g.kg-1 , from Guillaume et al., 1999).

Trout Carps Catfish Tilapias Eel Fish meal 300 250 80 150 650 Meat meal 40 Whey 100 Soybean cake 130 60 482 200 Maize gluten 170 50 Wheat germ 165 410 100 Maize 312 Wheat 200 Sorghum 450 Rice bran 55 Gelatinized / Extruded starch 210 Fish oil 115 15 50 Vegetal oil 50 20 Suet 50 Vitamin mixture 10 5 0.5 20 Mineral mixture 10 5 0.5 20 Bi-calcium phosphate 25 10 Binder 30 Crude Proteins (g.kg-1 MS) 380 310 320 290 360–400 Crude Lipids (g.kg-1 MS) 150 125 40 80 120–150 Digestible energy (MJ.kg-1 MS) 17.2 16.4 13.0 14.5 18.0

1

Table 9. Source of fish (3000 Cyprinus carpio.ha-1, 1000 silver carp.ha-1, 450 grass carp.ha-1 and 3500–7000 tilapia.ha-1) and prawn (5000–15 000 Macrobrachium rosenbergii.ha-1) growth in

polyculture ponds as indicated by 12C/13C analyses in a pond receiving 25 % protein content pellets at a daily rate of 2% of the tilapia biomass plus 6% of the common carp biomass reducing to 4% of the carp biomass when carps weight exceeded 150 g. (From Schroeder G.L. (1983). Sources of fish and

prawn growth in polyculture ponds as indicated by δC analysis. Aquaculture 35, pps. 29–42).

Percentage of growth originating from: Feed pellets Bottom, rock slime

and micro-algae Common carp < 50 > 50

Tilapia hybrid < 40 > 60

Probably < 20 % Probably > 80 %

Silver carp 0 100

Freshwater prawn 0 100

1

Table 10. Relative contributions of the autotrophic and heterotrophic food chains to the growth of fish (3000 Cyprinus carpio.ha-1, 1000 silver carp.ha-1, 450 grass carp.ha-1 and 3500-7000 tilapia.ha-1) and freshwater shrimps (5000–15000 Macrobrachium rosenbergii.ha-1) in ground ponds receiving a high level of organic fertilization : 50–200 kg.ha-1.day-1. (From Schroeder G.L. (1983). Sources of fish and

prawn growth in polyculture ponds as indicated by δC analysis. Aquaculture 35, pps. 29–42).

Percentage of growth stemming from: Heterotrophic

pathway Autotrophic

pathway Fish

Carpe commune 50–65 50–35

Hybrid tilapia 20–40 80-60

Silver carp 0 100

Freshwater shrimps

Macro brachium rosenbergii 30–50 50–70

1

Table 11. Chemical composition of some organic manures. (from Delincé G. (1992) The Ecology Of The Fishpond Ecosystem, With Special Reference To Africa,

230 pps. Dordrecht, Netherlands : Kluwer Academic).

Animal Ratio Feces: Urine Humidity Nitrogen

(%) Phosphorus

(% P2O5) Potassium (% K2O)

Milky bovine 75:25 85 4 2.7 3.4

Meat bovine 75:25 85 7.8 2.5 3.6 Chicken 100 65 6 7.9 3.7 Pig 53:47 85 4.7 4.3 2.7 Sheep 67:33 70 3.9 2.4 3.5 Horse 75:25 75 2.3 1.3 1.4

1

Table 12. Some fish-animal associations and related fish production (t.ha-1.year-1).

Pays Number of animal.ha-1 Fish production Duck

India 700 4.5 China 2500 3.4 Israel 900–2000 6.6–7.4 Africa 1000–1500 3.8–4.5 Hungary 300-500 3–3.5 Philippines 750–1250 2.2–4.5 Java 800–2000 0.7–1.3 Taiwan 2200 5.6 Taiwan 1500 3.5 Pig Africa 100 5.2–10.1 China 30–45 2–3 Brazil 25 2.2 Africa 50-100 7.7–10.2 Africa 30 4.8 USA 25–225 0.6–1.8 USA 40–70 3.2–4.1 USA 60 2.3–2.7 Hungary 40–80 2.7 India 80 7.3 Philippines 40–140 4.7–8.7 Chicken Africa 1250 2 Africa 2500 5.4 South Africa 6000 6.3

1

Table 13. Typology of fishfarming polycultures practiced around the world (From Billard R. (1980). La Pisciculture En Etang, 434 pps. Paris, France : INRA, modified).

I. Traditional associations in Europe 1. The same fish, at different ages 2. Several fish: carps, tench, roach, and pike etc. 3. Combination of 1 and 2. 4. Improvement of the traditional associations: pikeperch, catfish, and carassin etc. 5. Association in time:

• One species added during one season (coregone put in carp pond during winter) • Succession of two species (salmonids during winter, thermophilous fishes during

summer II. Traditional associations in Asia (based on phytophagous fish association)

1. Dominance of phytoplanktivorous fishes 2. Dominance of herbivorous fishes 3. Dominance of molluskophagous fishes 4. More complex associations, with increasing number of species

1

Table 14. Stocking density of some polyculture ponds, as a percentage of the different fish species (From Billard R. (1980). La Pisciculture En Etang, 434 pps. Paris, France : INRA, modified).

Food type of main species Macrophyte Mollusk Phytoplankton Hypophthalmichthys molitrix 16 12 65

Aristichthys nobilis 10 7.4 10

Ctenopharyngodon idella 55 24.2 12

Cyprinus carpio

Cirrhinus molitorella 3 6

Mylopharyngodon piceus 42

Divers 19 10 8

1

Table 15. Simple annual cost and return for a 1.000 square meter rice-fish culture project (two croppings) in the Philippines as of 1990 (in : P, 1uS$ # 30 P)

Item Amount (P) 1. Costs a) Variable costs Seeds at P6/kg 60 Fingerlings, 1000 pcs O. niloticus (15 g each) at P0.25 each 250 Fertilizers 305 Pesticides/chemicals 245 Labor 1620 Screen 30 Sub-total 2510 b) Fixed costs Interest on loan 90 Land amortization 116 Taxes 57 Irrigation Fee 88 Sub-total 351 Total costs 2861 2. Returns 3 750 900 Threshed palay at P250/cavan of 50 kg Marketable fish at P30/kg 4 650 Total returns 1 789 3. Net Income

1

Table 16. Semi - annual budget for a 1.000 square meter excavated earthen fishpond integrated with a head pig fattening, as of 1990 (in P, 1 US$ # 30 P)

Item Amount (P) 1. Capital Investment

Fishpond construction 7 500 (Excavation including water control structures)

Pig house 9000 Total Capital Investment 16 500

2. Costs a) Variable costs

Weanlings, 10-12 kg each at P800/head 8000 Fingerlings 4000 pcs at P0.20 each 800 Feeds 13 320 Labor 5520 Drugs and medicine 666 Repairs and maintenance 25 Miscellaneous 500

Sub total 29 631 b) Fixed costs

Pig house depreciation 675 Total costs 30 306 3. Revenues

Finished hogs 28 050 Fish Tilapia 10 140 Carp 1575 Total revenues 39 765

4. Net Income 9459

1

Table 17. Annual budget for a tilapia cage culture farm in the Philippines.

1. Investment cost (in P) 10 cages including nets and caretaker's house 300 000 Weighing scale 8000 Plastic boxes 20 000 Transport equipment 300 000 Total 628 000 2. Production cost Cost/kg Fry 120 000 fry at PO.25/pc P30 000 1.70 Feed at P10/kg: FCR = 1:2 300 000 11.90 Labor Caretaker 25 000 0.99 Helper (cleaning & inspection of cages) 15 750 0.62 Harvesting (casuals) 10 500 0.42 Driver 15 750 0.62 Ice 12 500 0.50 Depreciation (economic life = 5 years) 65 600 2.60 Amortization (cost of cage is amortized for 2 years)

150 000 5.95

Fuel and oil 15 000 0.60 Total P640 000 P25.40 3. Sales 25 200 kg × P35/kg P882 000 Less: Production cost 640 100 4. Net Profit P 242 000

Production data Stocking rate = 12.000 fry/cage Survival rate = 70 % Average final weight = 300 g Production 2 520 kg/cage × 10 cages = 25 200 kg

1

Table 18. Rates of returns of different types of farming in the Philippines as of 1990.

System % Rate of return Rice 33 Fish 48 Rice-Fish 42 Livestock 53 Fish-Livestock 63

1

Table 19. Financial analysis for a 20 m3 tilapia culture floating cage in Niger (West Africa) as of 1995 (in FCFA, 1 US$ # 600 FCFA).

Items (F CFA) Per 20 m3 cage

Per kg of fish

Costs (%)

Costs Fixed costs 29 500 675 11.0

Amortization of cage (7 yrs) 28 000 640 10.5 Amortization of small equipment (3 yrs) 1500 35 0.5

Variable costs 243 000 5585 89.0 Cage maintenancea 5000 115 2.0 Fry: 2.200 × 45 CFAb 99 000 2275 36.0 Feed: 363 × 3 x 100 CFAc 109 000 2505 40.0 Transport (fry, feed, marketable fish) 30 000 690 11.0

Total costs 272 500 6260 100.0 Gross sale: 435 kg × 850 F CFAd 369 750 8500 Net profit 97 250 2240 a The cost for maintenance takes only into account the materials needed for repairing the cages (plastic

net and floating structure), excluding the labor. b Number of pieces for stocking a grow-out cage x unit price of fingerlings %. c Net yield (kg) x feed conversion ratio x price of 1 kg of compounded feed. d Gross yield x price of 1 kg of market-size fish.

1

Table 20. Economic efficiency ratios for different fish culture models developed in Côte d'Ivoire (West Africa) (in FCFA, 1 US$ # 600 FCFA).

Models Ratios Average farm NP/T 14 189 DWR 2 742 M1 RRC 47% NP/T 13 604 DWR 2 501 M2 RRC 43% NP/T 22 009 DWR 3.919 M3 RRC 52%

NP/T: Net profit per 100 m² of pond per year in F CFA. DWR: Daily work remuneration and family management in F CFA. RRC: Rate of return of capital. Model M1: Heavy extension structure, pilote fish farmers, monosex male tilapia, predators:

Clarias, balance diet (including fish meal) Model M2: Heavy extension structure, Piloted fish farmers, monosex male tilapia, and

predators: Clarias, rice bran. Model M3: Heavy extension structure, monosex male tilapia, predators (Hemichromis) +

polyculture with Heterotis niloticus and Heterobranchus isopterus integrated with pig rearing for organic fertilization and supplementary feed (rice bran).