3. FEEDING METHODS - FERTILIZATION AND SUPPLEMENTARY DIET FEEDING

3.1 Introduction

At present over 90 percent of finfish and shrimp aquaculture production within third world and developing countries (including Latin America and the Caribbean) is realized within semi-intensive or extensive pond production systems employing a fertilization and/or supplementary diet feeding strategy. Here, in contrast to complete diet feeding, the dietary nutrient requirements of the farmed species are met either entirely or partly (in conjunction with an exogenous supplementary diet) through the production and consumption of natural live food organisms within the water body in which the fish or shrimp are cultured.

3.2 Pond Fertilization

3.2.1 The pond ecosystem and primary nutrient cycles

Since the aim of a fertilization feeding strategy is to augment the production of natural food organisms within a water body, it is perhaps useful to first describe the basic aquatic food chain or ecosystem and the underlying primary nutrient cycles operating within a pond ecosystem. Figures 8 and 9 show a generalized model of a simple aquatic ecosystem and an example of a natural pond food web ending in common carp (C. carpio), respectively. All aquatic ecosystems, including a fertilized fish or shrimp pond, rely on the simultaneous operation of two interlinked food chains; a light dependent “autotrophic” and grazing food chain, and a non-light dependent “heterotrophic” or detritus food chain. As the name suggests, the autotrophic or organic matter synthesizing food chain relies on the fixation of solar energy by green plants during photosynthesis with the production of new organic matter from carbon dioxide and water, and the subsequent consumption of these plant organisms by grazing animals. Although green plants, and in particular phytoplankton are the principal autotrophs or “primary producers” operating within a pond ecosystem, certain non-photosynthetic anaerobic bacteria and blue-green algae are autotrophic in that they are able to synthesize organic matter (ie. new cell biomass) from inorganic carbon by using chemical energy derived from the cellular oxidation of inorganic substrates such as hydrogen sulphide, sulphur, nitrogen, divalent iron and hydrogen (collectively these are termed chemosynthetic autotrophs as opposed to the photosynthetic autotrophs). By contrast, the heterotrophic or organic matter consuming food chain relies on the microbiological degradation of non-living organic matter or detritus into new microbial biomass with the release of inorganic nutrients and carbon dioxide; the new microbial biomass (mainly bacteria) serving as a feed source for protozoa, nematodes and other benthic animals, and the released inorganic nutrients and carbon dioxide in turn being available for further photosynthetic production by the primary producers or autotrophs.

All pond food organisms, including autotrophs and heterotrophs, consist mainly of carbon-C, nitrogen-N and phosphorus-P (ie. composition of phytoplankton grown on a nutrient rich medium being about 45–50%C, 8–10%N and 1%P on a dry basis: Edwards, 1982), and consequently are dependent on the biological supply of these primary nutrients for their growth. The basic chemical and biological pathways involved in the supply and cycling of C, N and P within a natural pond ecosystem are shown in Figure 10, 11 and 12, respectively. From an understanding of these nutrient cycles it can be seen that the natural productivity of an enclosed water body can be increased with careful management, through the controlled addition of chemical inorganic fertilizers (by feeding the autotrophic food chain) and/or organic manures (by feeding the heterotrophic food chain).

Figure 8. Generalised representation of a simple aquatic ecosystem. The lightly shaded blocks represent the biomass of each type of organism. The stippled arrows show the direction and magnitude of energy flow while the single line arrows indicate the transference of nutrients either through direct consumption, excretion or death, and bacterial decay. Energy is the amount of solar energy taken up by the primary producers (green algae and higher plants). Ecosystems are usually divided into a grazing food chain of large animals and a detritus or decomposer food chain of microorganisms (Source: Eltingham, 1971).

Figure 9. Schematic representation of a pond food web ending in common carp (Cyprinus carpio; Hepher and Pruginin, 1981)

3.2.2 Preparation of the pond bottom prior to fertilization

The soil of the pond bottom, and in particular the mud layer ¹, is considered to be the “chemical laboratory” and “primary nutrient store” of the pond ecosystem, and as such plays a vital role in the maintenance of pond productivity (Figure 10–12: Mortimer, 1954; Huet, 1975; Vincke, 1985; White, 1986). However, the success of a pond fertilization feeding strategy, in many instances, depends upon the initial drying and/or chemical treatment of the pond bottom with lime.

3.2.2.1 Pond drying

The advantages of air drying and exposing the pond bottom to atmospheric oxygen and sunlight prior to fertilizer application have been summarised by Mortimer (1954), Vincke (1985), Clifford (1985), Fast (1986), Stokes and Smith (1987) and Wilson (1987), and include:

• Improvement in soil texture and primary nutrient availability for future phytoplankton production by facilitating the breakdown and decomposition of organic matter through oxidation with consequent mineralization of the surface mud layer.

• Reduced mud sediment demand for oxygen once the pond is filled with water.

• A well aerated and partially oxidised soil, makes the bottom better suited for colonization by desired benthic food organisms.

• Oxidation and removal of undesirable metabolites, such as hydrogen sulphide (a by-product of anaerobic respiration of sulphur bacteria), which if allowed to accumulate may inhibit the growth of phytoplankton and the cultured fish or shrimp.

• Elimination of fish or shrimp predators, parasites and their eggs, and unwanted aquatic macrophytes.

• Facilitates the cropping of the cultured fish or shrimp and the removal of excessive mud or silt deposits from the pond bottom, the latter being a valuable fertilizer for agricultural crops.

¹ The pond mud or sediment generally consists of a mixture of settled organic matter or detritus (dead plant/animal fragments and faecal matter; fresh or in a state of bacterial/microbial colonization and decomposition), live benthic organisms (algae, protozoa, nematodes oligochaetes, polychaetes, gastropods and insect larvae), and inorganic minerals. The latter may be present as coarse sand or silt particles, precipitated mineral salts, bound cations adsorbed onto negatively charged colloidal clay/humus particles, or as free dissociated cations within the interstitial water of the pond mud (Boyd, 1982; Coche, 1985).

Figure 10. The carbon cycle

¹ Equilibrium depends on pH; the solubility of CO₂ increasing with pH. In addition to the inorganic C forms shown, precipitation of calcium carbonate may occur from the bicarbonate ( Ca(HCO₃)₂ ⇌ CaCO₃ + H₂O + CO₂). Particulate and colloidal calcium carbonate plays an important role in that it has the capacity to strongly adsorb a variety of biologically active compounds, including humic acids and phosphates.

Figure 11. The nitrogen cycle

Figure 12. The phosphorus cycle

¹ Equilibrium depends on pH; the solubility of orthophosphate acid increasing with pH.

² Slow release of orthophosphate from pond sediments, particularly under reducing conditions (caused indirectly through metabolism of anaerobic sulphur bacteria).

The drying out period for adequate mud mineralization is usually between five to ten days, as evident by the appearance of cracks on the mud surface or by the ability of the pond bottom to support a man's weight without subsiding (Vincke, 1985; Clifford, 1985; Wilson, 1987). For the culture of specific benthic food organisms, it is essential that the pond bottom is not ‘bone’ dry; for example a drying period 7–10 days and 3 days is usually recommended for the preparation of pond muds for the growth of “lab-lab” (algal mat primarily composed of blue green algae and diatoms) and “lumut” (algae mat primarily composed of filamentous grass-green algae) within brackishwater fish or shrimp ponds, respectively (ASEAN, 1978). Although ponds are usually dried at the start of each new culture cycle, in China fish ponds are normally dried for a 15–20 day period only every one to three years (FAO, 1983). However, pond drying is not normally recommended for those coastal and riverplain soils such as “cat's clay” and “mine overburn” which contain pyrite - FeS₂ and other sulphur containing minerals. Upon exposure to air these minerals oxidize to form sulphuric acid and iron sulphate compounds (jarosite); the resultant “acid-sulphate” soil is characterized by a very low pH (< 4) and yellow spots or streaks of jarosite (Coche, 1985). Other disadvantages often ascribed to pond drying include 1) loss in time otherwise used for fish or shrimp production, and 2) additional labour and water cost (ie. cost of draining and refilling the pond with water, including electrical pumping costs).

3.2.2.2 Liming

According to Thomaston and Zeller (1961) and Boyd (1986), for a freshwater pond to respond properly to fertilization, the bottom mud must not be highly acidic and the surface water should have a neutral-alkaline pH (7–8) and a total alkalinity and total hardness of 20 mg/l or more as calcium carbonate. Acidic muds strongly adsorb inorganic phosphates, and pond food organisms (particularly phytoplankton) do not grow wellin an acidic environment (pH5–6) or in water with a low base carbon and calcium concentration (Miller, 1976; Vincke, 1985; Fast, 1986; Boyd, 1986). However, these imbalances may be corrected by applying quicklime (CaO) or limestone (CaCO₃) to the pond bottom or water column prior to the start of the culture cycle or pond fertilization programme. Boyd (1982) lists three basic types of ponds that respond favourably to liming: 1) dystrophic pondswith waters heavily stained with humic substances and muds with large stores of slowly decaying organic matter (typical water quality: pH 5 – 6, alkalinity 1 – 5 mg/l CaCO₃, acidity 0 mg/l CaCO₃), 2) ponds with waters of low pH and alkalinity because of moderately acid muds and watershed soils (typical water quality: pH 5.5–7, alkalinity 3 – 15 mg/l CaCO₃, acidity 0 mg/l CaCO₃), and 3) dystrophic ponds with waters containing mineral acidity resulting from acid-sulphate soils of watersheds (typical water quality: pH 2 – 4.5, alkalinity 0 mg/l CaCO₃, acidity 10 – 250 mg/l CaCO₃).

The beneficial effects of liming ponds can be summarised as follows:

• Liming raises the pH and alkalinity of acid waters to desirable levels, establishing a alkaline reserve or pH buffering system (Mortimer, 1954; Huet, 1975; Miller, 1976; Boyd, 1982, FAO, 1983; Yamada, 1986).

• By virtue of its effect on alkalinity, liming increases the availability of carbon for photosynthesis (Mortimer, 1954; Miller, 1976; Boyd, 1982; Fast, 1986; Yamada, 1986).

• Liming raises the pH of the mud bottom to desirable levels and consequently reduces the capacity of the mud to adsorb plant nutrients such as inorganic phosphates, thus increasing their bioavailabilityto pond food organisms (Mortimer, 1954; Huet, 1975; Miller, 1976; Boyd, 1982; FAO, 1983; Fast, 1986; Yamada, 1986).

• By raising the pH of acidic sediments, liming creates a more favourable environment for microbial growth and consequently accelerates the decomposition and mineralization of organic matter within the sediment (Huet, 1975; Miller, 1976; Boyd, 1982; Fast, 1986; Yamada, 1986).

• By raising the alkalinity and hardness of water; liming serves as direct source of soluble calcium for pond food organisms (Mortimer, 1954; Boyd, 1982; FAO, 1983).

• Liming assists in the clarification of turbid waters by facilitating the flocculation and precipitation of organic/clay colloids in suspension (including humic acids), there-by improving light penetration for photosynthesis (Mortimer, 1954; Boyd and Scarsbrook, 1974; FAO, 1983; Yamada, 1986).

• Liming serves as a pond disinfectant, by killing fish parasites and their intermediate hosts, animal competitors, and unwanted green plants (Mortimer, 1954; Huet, 1975; Miller, 1976; Boyd, 1982; FAO, 1983; Yamada, 1986).

• By virtue of the above attributes, liming therefore, increases (all be it indirectly) the natural productive capacity of freshwater ponds with acidic waters and low total alkalinity and hardness (Boyd, 1982; Strumer, 1987). For example, Zeller and Montgomery (1962) and Boyd and Scarsbrook (1974) reported increased live food production (phytoplankton, zooplankton and benthic food organims) within fertilized ponds which were pretreated with lime; liming increasing the effectiveness of the fertilization strategy employed. A similar relationship exists for fish production within such ponds; pretreating ponds with lime increasing the effectiveness of inorganic fertilization with increased fish yields. For example, Arce and Boyd (1975) and Hickling (1962) reported a 24.9% and 41.6% increase in Tilapia production, respectively, within limed over unlimed ponds receiving inorganic fertilizer inputs. However, it should also be mentioned that not all studies have demonstrated a positive influence of liming on fish production (Swingle, 1947; Miller, 1976; Boyd, 1982). Thus on the basis of liming trials conducted within freshwater fish ponds in Africa, Miller (1976) concludes that the use of lime in water of pH greater than 6.5 appears unnecessary. Similarly, with the exception of coastal acid-sulphate soils, liming is not considered to be essential for saltwater pond preparation; seawater exhibiting a strong buffering capacity and having a pH range between 7.5 to 8.4 (Sturmer, 1987).

According to Boyd (1982), the lime requirement for a fish pond should represent the quantity of calcium carbonate required to raise the pH of the mud to 5.9 so that the base unsaturation (proportion of acidic cations to total cations on particle exchanges sites) of the mud will be 0.2 or less and the total hardness (and alkalinity) will be above 20mg/l. Table 12 shows the recommended lime application rates for fish ponds as determined by the Boyd technique. However, it should be remembered that the above relationship between base unsaturation and pond alkalinity/hardness was determined for ponds in Alabama, USA and relationships between mud pH and base unsaturation differ geographically (Boyd, 1986).

Table 12. Estimated lime requirement (kg CaCO₃/ha) needed to increase the total hardness and alkalinity of pond water to 20mg/l or greater¹

¹ Source: Boyd (1982)

² Lime required (as CaCO₃) is estimated from the pH of the pond muds beforeand after the addition of a buffer solution. The mud sample for limerequirement measurement should be dried at room temperature by spreadingin a thin layeron a plastic sheet. The dried mud sample is then groundusing a pestle and mortar and passed through a 20-mesh sieve (0.85mmopenings) for pH analysis. The buffer solution is prepared by dissolving20g of p-nitrophenol, 15g of boric acid, 74g of potassium chloride, and10.5g of potassium hydroxide in distilled water and diluting to one litrein a volumetric flask. Place 20.0g of the dried and ground mud sample intoa 100 ml beaker, adding 20ml of distilled water, and stir intermittently for one hour. Measure the pH of the mud-water mixture with a glasselectrode while stirring. The value obtained is the mud pH. Next,add 20.0ml of the prepared buffer solution to the mud-distilled watermixture and stir intermittently for 20 minutes. Set the pH meter atpH 8.0 with a 1:1 mixture of buffered solution and distilled water,and then determine the pH of the mud-distilled water-buffer solutionmixture while stirring vigorously. If the pH of the mud-distilledwater-buffer solution mixture is below 7.0, repeat the analysis with10.0g dry mud and double the liming rate from the Table above (for adetailed description of the method see Boyd, 1979).

Ideally, the relationship between pH and base unsaturation of muds should be determined for every farm or region, and liming rates computed accordingly. A simple method for determining the lime requirement of pond muds that does not require data on the relationship between pH and base unsaturation has been developed by Pillai and Boyd (1985); the liming rate (kg CaCO₃/ha) is simply determined by measuring the pH change in 40ml of buffer solution (10g p-nitrophenol, 7.5g boric acid, 37g potassium chloride, and 5.25g potassium hydroxide dissolved and diluted to 1000ml with distilled water; the buffer pH being adjusted to 8.00) caused by adding 20g of ground dried mud (particles < 0.85mm) and multiplying the observed pH change by 5600.

The above techniques for estimating lime application rates do not apply to acid sulphate soils, since these sediments have both exchange and sulphuric acid acidity. Singh (1980) recommends a soil tilling (pyrite oxidation) and leaching reclamation procedure, followed by liming and inorganic/manure fertilization, for the management of acid sulphate soils. A procedure for estimating the lime requirement of acid-sulphate soils is given by Boyd (1979). Examples of lime application rates for aquaculture ponds suggested by other workers are shown in Table 13.

For the composition and neutralizing value of commonly used liming materials see Boyd (1979) and Tacon (1987a). Although the neutralising effects of quicklime (CaO) and slaked lime (Ca(OH)₂) on acid waters is higher and faster than that of agricultural limestone (CaCO₃), the latter is generally regarded to be the safest, cheapest and most effective liming material for ponds (Boyd, 1982). On a general basis, liming materials should be added 2 – 3 weeks prior to fertilization (Boyd, 1982; Miller, 1976), and applied by spreading evenly over the pond bottom (in the case of empty ponds) or water surface. The residual effects of liming depend on water exchange within the pond, and may last for several years if water exchange is not excessive. For example, Boyd (1979) found that an annual lime application rate of 25 percent of the initial dose of about 4400kg/ha agricultural limestone was sufficient to maintain adequate water quality and mud pH for an eight year period within annually drained fish ponds in Alabama, USA.

Table 13. Examples of suggested lime application rates for aquaculture ponds

1. Suggested liming rates for the treatment of low soil pH ¹

   Soil pH	Liming material (lbs/acre) 2
   	Carbonate of lime	Slaked lime	Caustic lime
   4	1487	1417	994
   4.5	1320	1258	898
   5	994	924	634
   5.5	660	634	466
   6	334	299	239
   6.5	0	0	0

   ¹ Source: Clifford (1985)
   ² 1kg = 2.205 lbs, 1 ha = 2.47 acres

2. Suggested liming rates for pond muds based on pH and texture of muds ¹ Lime requirement (kg/ha of CaCO3)

   Mud pH	Heavy loams or clays	Sandy loam	Sand
   4	14,320	7,160	4,475
   4 – 45	10,740	5,370	4,475
   4.6 – 5	8,950	4,475	3,580
   5.1 – 5.5	5,370	3,580	1,790
   5.6 – 5.0	3,580	1,790	895
   6.1– 6.5	1,790	1,790	0

   ¹ Source: Schaeperclaus, 1933 (cited by Boyd, 1982)

3. Suggested liming guide for aquaculture ponds in Rwanda (Schmidt and Vincke, 1981)

   Newly constructed ponds with acid water (pH 4–6.5): use of powdered agricultural limestone at a rate of 1500–2000kg/ha by spreading on the dried pond bottom and lightly tilling the lime into the mud surface layer, then filling pond with water.
   Other ponds :monthly application of powdered limestone at a rate of 150–200kg/ha.
   For a review on the use of lime in African countries see Miller (1976).

4. Suggested liming guide for aquaculture ponds - general (Huet, 1975)

   Liming pondwater: the use of up to 200 kg/ha/day of quicklime (CaO). Liming pond bottom to control parasites: the use of 1000–1500kg/ha of quicklime (CaO) or 1000kg/ha of calcium cyanamide. Liming materials should be spread on pond bottom which is still damp.

   Liming pond bottom to improve the mud before using other fertilizers: the use of 200–400kg/ha of quicklime (CaO) provided that the pond is not acid. If the aim of liming is to increase the pH and alkalinity of an acid pond, in principle 200kg/ha of quicklime (CaO) is generally sufficient to raise the alkalinity by one unit.

5. Suggested liming guide for African catfish ponds (Viveen et. al., 1985)

   Newly constructed ponds: use of agricultural lime at a rate of 200– 1500kg/ha and mixing with the upper layer (5cm) of the dried pond bottom. Pond is then filled with water (till 30cm) and left for one week prior to fertilization.

   Used ponds: use of 100–150kg/ha quicklime (CaO) added to damp pond bottom to eliminate pathogens, parasites and invertebrate predators. Pond is then left for a 7–14day period and then filled with water to a depth of 30cm, and pH of water adjusted by adding agricultural lime.

6. Liming rates employed for aquaculture ponds in China (FAO, 1983)

   Liming pond water: use of quicklime (CaO) at a rate of 750–900kg/ha and 900–1125kg/ha for ponds containing 6–7cm water and containing little silt and silt respectively. For ponds containing a considerable amount of water (unspecified) an application rate of 1875–2250kg/ha/ month quicklime is used.

7. Liming rates suggested for car ponds in Hungary (Horvath, Tamás and Tölg, 1984; ADCP, 1984)

   Nursery ponds: use of 200–500kg/ha of lime (CaO) on dried bottom for disinfection, followed by aeration of pond bottom by tilling.

8. Liming rates suggested for Colossoma sp in Brazil (Woynarovich, 1986)

   Nursery ponds: use of 150–300kg/ha of limestone (CaCO³) on dried bottom.

9. Liming rates suggested for Macrobrachium rosenbergii in Panama (MIDA, 1984)

   Newly constructed ponds: use of 500–1000kg/ha of limestone (CaCO₃) on pond bottom.

10. Liming rates suggested for newly constructed rural fish ponds in Thailand (Edwards and Kaewpaitoon, 1984)

    Acidity of new pond is tested using litmus paper after introduction of water to a depth of 10cm.
    For water pH 4.5–6: 500kg quicklime/ha
    For water pH 4–4.5: 1250kg quicklime/ha. After one day, the pond should be filled with water and the acidity checked again.

3.2.3 Chemical fertilization of aquaculture ponds

Chemical fertilizers are applied mainly to increase the primary productivity of aquaculture ponds. The chemical nomenclature and composition of the major single and multinutrient chemical fertilizers used in aquaculture has been presented previously (Section 1.2 and 3.12; Tacon, 1987a).

3.2.3.1 Effect on pond productivity and fish/shrimp production

Chemical fertilizers act principally on the autotrophic and grazing food chain by directly stimulating phytoplankton production within the pond (Hepher, 1962; McIntire and Bond, 1965; Hall, Cooper and Werner, 1970; Djajadiredji and Natawiria, 1965; Boyd, 1973; Miller, 1976; Guerrero and Guerrero, 1976; Cruz and Laudencia, 1980; Davidson and Boyd, 1981; Hepher and Pruginin, 1981; Bishara, 1978; Rubright et. al., 1981; Nailon, 1985; Olah et. al., 1986; Pruder, 1986; King and Garling, 1986; Yamada, 1986). For example, the studies of Hepher (1962) showed that the production of phytoplankton within chemically fertilized fish ponds in Israel was four to five times higher than equivalent ponds receiving no fertilizer input; the primary productivity of chemically fertilized ponds ranging from a carbon uptake of 4–8g/m²/day during the summer (mid-day water temperature 25–30°C) to 2.5–5g/m²/day during spring and autumn (mid-day water temperature 20–25°C). According to Schroeder (1978) over 90% of the total primary production is smaller than 40 microns in size. As a consequence of their direct effect on phytoplankton production, chemical fertilizers also indirectly augment the production of grazing zooplankton (McIntire and Bond, 1962; Dendy et. al., 1968; Hall, Cooper and Werner, 1970; Lyubimova, 1974; Rubright et. al., 1981; Torrans, 1986) and benthic food organisms (Ball, 1949; McIntire and Bond, 1962; Sumawidjaja, 1966; Rubright et. al., 1981; Boyd, 1981). For example, Torrans (1986) reports a standard zooplankton biomass range of 2–10g/m³ within inorganically fertilized static fish ponds.

Although aquaculture production within chemically fertilized ponds will vary depending on the feeding habit and density of the culture species stocked, considerable increases in fish and shrimp production are possible (Table 14). For example, Schroeder (1978) reports that the maximum fish yields attainable with no supplementary feeding (in earthen ponds in Israel) are 1–5kg/ha/day and 10–15kg/ha/day for ponds receiving no fertilizer input and chemical fertilizers, respectively; common carp, tilapia and silver carp polyculture at 4500–9500 fish per hectare. Similarly, Horvath, Tamas and Tolg (1984) report a fish production increase (mainly carp polyculture) within earthen ponds in Hungary of 11–25kg and 15–30kg from a 200kg fertilizer input of superphosphate or ammonium nitrate, respectively. However, as mentioned previously, the success of a chemical fertilization strategy will depend upon the ability of the farmed fish or shrimp species to take advantage of the increased primary productivity within the pond. Adult fish and shrimp species which can feed directly on primary autotrophs, include: Phytoplankton - Silver carp (Hypophthalmichthys molitrix), Indian carp (Catla catla), Tilapia (esculentus, aureus, niloticus, kottae, mariae, galilaeus, leucostictus, mossambicus), Bighead carp (Aristichthys nobilis); Benthic algae - Milkfish (Chanos chanos), Tilapia (mossambicus, guineensis, melanotheron, niloticus), Mullet (Mugil cephalus), Rabbit fish (Siganus sp.) Rohu (Labeo Rohita), Freshwater prawn (Macrobrachium dayanum, M. lanchesteri), Metapenaeid shrimp (Metapenaeus ensis M. affinis, M. macleay), Penaeid shrimp (Penaeus vannamei); Vascular aquatic plants - Grass carp (Ctenopharyn-godon idella), Tilapia (rendalli, niloticus, Mossambicus, zillii), Wuchang fish (Megalobrama amblyocephala), Rabbit fish (Siganus sp.) Rohu (Labeo rohita) and occasionally freshwater prawns (Macrobrachium sp.). For a review of the natural feeding habits of the major cultivated fish and shrimp species see Ling (1969), Wickins (1976), Von Westernhagen (1974), Bowen (1982), Cremer and Smitherman (1980), Bishara (1979), Cruz and Laudencia (1980), Guerrero and Guerrero (1976), Rubright et.al., (1981), Weidenbach (1982), Swift (1985), Horvath, Tamas and Tolg (1984), Torrans (1986), King and Garling (1986), New (1987), Hunter, Pruder and Wyban (1987), and Lilyestrom and Romaire (1987).

Table 14. Reported fish and shrimp production increases within chemically fertilized ponds compared with non-fertilized control ponds

1 Cited by Hepher and Pruginin (1981)

3.2.3.2 Fertilizer application rates

It is generally accepted that inorganic phosphate-P and nitrogen-N are the two major soluble nutrients normally limiting the algal productivity of aquaculture ponds; phosphate-P and nitrogen-N generally being the first limiting nutrients (ie. most essential from a pond fertilization viewpoint) within freshwater and brackishwater ponds, respectively (Boyd 1982, 1986; Miller, 1976; ASEAN, 1978; Vincke, 1985; Smith, 1984; Nailon, 1985; Yamada, 1986; Strumer, 1987; Boyd and Minton, 1987). It must be emphasized at the outset that no two ponds are alike and that a fertilization programme developed or recommended for one location may be totally unsuitable for another, the response of a fertilization programme depending on the pond's morphology, hydrology, environment, bottom sediment, and water quality, on the aquaculture species cultured and the fertilizer used, and the fertilizer application method and rate employed (Yamada, 1986; Boyd, 1986). Clearly, every farm must be considered as being unique, and a personalised fertilization programme developed accordingly. However, despite this rather daunting picture, some generalisations can be made regarding pond fertilization.

According to Hepher (1963, 1967) there is no biological or economical justification of applying higher fertilizer dosages than 0.5mg phosphate-P/l or 1.4mg nitrogen-N/l for freshwater ponds in Israel; applications higher than these levels generally being fixed as precipitated phosphates or lost to the environment as gaseous ammonia. The above levels are equivalent to a fertilizer application rate of 60kg/ ha single superphosphate (11kg P₂O₅/ha) and 60kg/ha ammonium sulphate (13kg N/ha) applied at 2-weekly intervals (0.8–1.0m water depth, 8–10,000 m³ water/ha). This fertilizer application rate is currently the standard dose for fertilizing semi-intensive ponds in Israel with densities of 2000–3000 fish/ha (Hepher and Pruginin, 1981). Boyd (1982) and ASEAN (1978) suggest chemical fertilization strategies to maintain soluble nitrogen and orthophosphate at 0.1–0.5mg P/l (Boyd, 1982) and 0.95mg N/1 and 0.11mg P/l (ASEAN, 1978) within freshwater and brackishwater aquaculture ponds, respectively. The beneficial effect of using nitrogen based fertilizers within freshwater ponds has met with variable results (Hickling, 1962; Hepher, 1963; Miller, 1976; Boyd and Sowles, 1978; Boyd, 1982; Vincke, 1985; Yamada, 1986); Vincke (1985) suggest that the continued use of N-based fertilizers may not be necessary within tropical freshwater fish ponds due to the high rate of N fixation by free-living bacteria and blue-green algae within these ponds. Example of fertilizer application programmes which have been tested and proven under pond farming conditions are shown in Table 15.

Although various precise chemical methods exist for estimating the primary productivity of a water body (Boyd, 1979, 1982; Schroeder, 1978; Davidson and Boyd, 1981; Olah et.al., 1986), the effectiveness of a pond fertilization programme can be quickly determined by measuring the turbidity (ie. transparency) of the water body by means of a Secchi disk. This simple and practical method is based on the assumption that the main source of turbidity within a fish or shrimp pond is the abundance of phytoplankton (Barica, 1975; Almazan and Boyd, 1978; Boyd, 1979, 1982). Stickney (1979) and ASEAN (1978) recommend a Secchi disk visibility of 30cm to achieve and maintain proper fertilization; readings above (>35cm) and below (<25cm) this level indicating under and excessive phytoplankton production, respectively. If a Secchi disk is not available, the rule of thumb is to submerge one's arm to the elbow; if one is just able to see the ends of ones fingers the water should be productive enough (FAO, 1981). The Secchi disk method is not suitable for shallow brackishwater ponds intended for benthic algal production or for use within turbid water bodies containing high concentrations of suspended clay particles.

3.2.3.3 Factors influencing the action of chemical fertilizers

Apart from the beneficial effect of liming (section 3.2.2.2) the following factors are known to influence the success or not of a chemical fertilization feeding strategy;

1. Sunlight: In the presence of adequate inorganic nutrients, primary production reaches a maximum value set by the amount of solar energy penetrating the pond water (Schroeder, 1978, 1980; Wohlfarth and Schroeder, 1979). Although Tamiya (1957) and Hepher (1962) state that the maximum primary productivity within tropical waters is equivalent to about 10g of carbon fixed as algae/m²/day, Talling et.al., (1973) have suggested that the upper limit for gross primary productivity is 17.8g of carbon fixed/m²/day or the equivalent release of 47g of oxygen (in general 2.6g of oxygen are produced for every gram of carbon fixed during photosynthesis; Cassinelli et.al., 1979; Pruder, 1986). According to Pimentel and Pimentel (1979) about 0.03% of the light reaching an aquatic ecosystem is fixed by phyto-plankton and aquatic plants, and is calculated to be approximately 4 × 10⁶ kcal/ha/year or about one third of that fixed in terrestrial habitats.

From the above it follows that increasing water depth, water turbidity¹ (caused by suspended clay particles), over-cast skies and shading will reduce the amount of light reaching the green autotrophs, and consequently will limit the primary production capacity of a pond (Miller, 1975; Boyd, 1986). Furthermore, the continued application of chemical fertilizer beyond a certain level will not result in increased primary productivity; the amount of solar energy penetrating the pond water dictating the upper limit for autotrophic production (Hepher, 1962; Schroeder, 1978, 1980).

¹ The detrimental effect of water turbidity resulting from clay suspensions may be reduced by treating the pond water with aluminium sulphate or gypsum (Boyd, 1986), barnyard manure (2–3 applications of 1 ton/acre at 3-week intervals; Boyd and Snow, 1975), or a cotton-seed meal superphosphate mixture (3 : 1, 100 lbs/acre; Swingle and Smith, 1974).

Table 15 Examples of pond fertilization feeding strategies

FRESHWATER PONDS

1. Freshwater prawn (M. Lanchesteri/M. Lanceifrons montalbanense) and Tilapia (nilotica, mossambica) polyculture - Philippines (Guerrero and Guerrero, 1976):
   

   • Biweekly application of 50kg/ha ammonium phosphate 16:20:0 (NPK) using underwater platforms, water depth 0.6–0.7m

   • See also Guerrero (1981) for tilapia monoculture using 50kg/ha ammonium phosphate to maintain productivity, and stocking fish two weeks after initial fertilization.

2. Tilapia (nilotica) fingerling production - Rwanda (Schmidt and Vincke, 1981):
   

   • Monthly application of 40kg/ha superphosphate (18%) and 20–40 kg/ha ammonium sulphate or 10–20kg/ha urea and 100kg/ha agricultural limestone

3. Tilapia growout - Ivory Coast (Lazard, 1973; cited by Miller, 1976):
   

   • Monthly application of 60kg/ha triple superphosphate; monthly dose given in two equal applications in baskets suspended in the surface water

4. Carp (C. carpio) fingerling production - Malagasy Republic (Vincke, 1970; cited by Miller, 1976):
   

   • Biweekly application of 20–40kg/ha triple superphosphate and 40– 80kg/ha ammonium sulphate given every 2–3 weeks

5. Tilapia growout - Zambia (Strum 1966; cited by Miller, 1976):
   

   • Monthly application of 56kg/ha double superphosphate (38% P₂O₅)

6. Carp (C. carpio, H. molitrix) and Tilapia (aurea or hybrid) polyculture - Israel (Hepher, 1962):
   

   • Biweekly application of 60kg/ha superphosphate (18%) and 60kg ammonium sulphate

7. Carp nursery ponds - Hungary (Horvath, Tamas and Tölg, 1984):
   

   • After pond drying and disinfection with lime, pond half filled (5–7 days before fish stocking) with water and 150–200kg/ha ammonium nitrate or carbamide added. Half of the application to be added during pond filling and the remainder given in two applications after the first and second week of nursing. Phosphorus should be added in a semi-dissolved state at 100kg/ha when flooding of the pond occurs

8. General freshwater fish - Alabama USA (Boyd and Snow, 1975):
   
   • Eight to twelve periodic applications per year of 45kg/ha 20:20:5 (NPK, granular compound fertilizer), applied on to under-water platforms; first application, followed by two applications at biweekly intervals, three applications at triweekly intervals, and five applications at monthly intervals
   • Subsequent studies have shown that nitrogen fertilizers may be greatly reduced or omitted without reducing fish production (Boyd and Sowles, 1978)
   • Dobbins and Boyd (1976) also found that potassium fertilizers are generally unnecessary and that the standard phosphorus fertilization rate of 9kg P₂O₅/ha/application could be reduced by half without significantly decreasing fish production
   • two to three weekly application of 6.6kg/ha ammonium polyphos-phate solution (liquid fertilizer; 10:34:0 NPK; density 1.4g/ml; Davidson and Boyd, 1981)
9. Milkfish (C. chanos), Tilapia (T. nilotica) and Snakehead (O. striatus) polyculture - Philippines (Cruz and Laudencia, 1980):
   

   • Biweekly application of 50kg/ha ammonium phosphate (16:20:0, NPK)

10. General freshwater fish - China (FAO, 1983):
    

    • Application of compound fertilizer (4:4:2, NPK) to maintain concentration of 0.9, 0.9 and 0.45mg/l respectively

11. General freshwater fish - Brazil/Hungary (Woynarovich, 1985):
    

    • Biweekly application of 15kg/ha superphosphate (18%) and 30kg/ ha ammonium nitrate

BRACKISHWATER PONDS

12. Shrimp (Penaeus stylirostris) growout - USA (Rubright et.al., 1981):
    

    • Application of 15.7kg/ha pelletised urea (45:0:0) and 6.7kg/ha triple superphosphate 19 days before shrimp stocking, and again 7 days after stocking

13. Shrimp (Penaeus sp.) growout - Ecuador/Philippines (Clifford, 1985):
    

    • Quotes initial application (while filling the pond) of 16.8– 22.4kg/ha urea and 1.1–5.6kg/ha triple superphosphate, followed by weekly application of 2.2–5.6kg/ha urea and 1.1–2.2kg/ha triple superphosphate (application methods - platforms, suspended perforated bags or dumping into intake water)

14. Shrimp (Penaeus sp.) growout - USA (Colvin, 1985):
    

    • Quotes Parker et.al., (1974) who used weekly applications of 45kg/ha urea once the shrimp reached 20mm in length, together with a 25% protein pelleted shrimp diet

15. Shrimp (Penaeus sp.) growout - Ecuador (MIDA, 1985):
    

    • Quotes substitution of pelleted feeds in Ecuador with weekly applications of 20kg/ha urea and 7kg/ha P₂O₅, final shrimp density 1.5–2/m²

16. Shrimp (Penaeus sp.) growout - Mexico (unpublished data):
    

    • Initial application of 15–35kg/ha urea and 5–12kg/ha P₂O₅, followed by monthly applications depending on water productivity

17. Shrimp (Penaeus sp.) growout - Brazil (unpublished data on commercial sector):
    

    • Initial application of 10–150kg/ha urea (mean 25kg) and 5–60kg/ ha triple superphosphate (mean 15kg); followed by monthly applications depending on water productivity

18. Mullet (Mugil capito) growout - Egypt (Bishara, 1979):
    

    • Monthly applications of 20kg/ha superphosphate or 18kg/ha super-phosphate plus 18kg/ha ammonium nitrate

19. Red drum (S. ocellatus) nursery - USA (Colura, 1987):
    

    1. Fertilization schedule used at the GCCA/TPWD John Wilson Marine Fish Hatchery in Corpus Christi, Texas. All fertilizer rates are calculated on a per hectare basis:

       Day	Treatment
       1	Fill pond to 1/3 volume
       3	Add 12L phosphoric acid and 28L ammonium nitrate (33%N)
       6	Spread 455kg of cottonseed meal (CSM) over water surface
       8	Finish filling pond
       12	Add 12L phosphoric acid and 28L ammonium nitrate
       14	Stock approximately 750,000 fry
       16	Spread 114kg CSM over water surface
       22	Add 12L phosphoric acid and 28L ammonium nitrate
       24	Spread 114kg CSM over water surface
       30	Spread 114kg CSM over water surface
       38	Spread 114kg CSM over water surface

    2. Fertilization schedule used at the TPWD Perry R. Bass Marine Fisheries Research Station, Palacios, Texas. All fertilizer rates are calculated on a per hectare basis:

       Day	Treatment
       1	Spread 282kg CSM on dry pond bottom; fill to approximately 100cm deep
       3	Continue filling. Add 9L phosphoric acid and 4.6kg urea (45%N)
       7	Spread 31.3kg CSM
       10	Spread 31.3kg CSM, stock fry
       12	Spread 31.3kg CSM, add 3L phosphoric acid and 4.6kg urea
       15	Spread 31.3kg CSM
       17	Spread 31.3kg CSM
       19	Spread 31.3kg CSM, add 3L phosphoric acid and 4.6kg urea
       21	Spread 31.3kg CSM
       23	Spread 31.3kg CSM
       24	5.7kg/ha salmon starter diet
       25	Spread 31.3kg CSM, add 3L phosphoric acid and 4.6kg urea

    
    

20. ‘Lab-Lab’ (benthic blue-green algal complex) culture for milkfish (C. chanos)/shrimp ponds - General (ASEAN, 1978):
    

    • Pond bottom should be first dried for a 7–10 day period (not bone dry). Dried chicken manure (insecticide free) should be applied to dry pond bottom at the rate of 350kg/ha to increase the organic matter content of pond sediment; lab-lab growth being directly related to the organic matter content of the mud - growth being very abundant in soils with 16% organic matter or more. In the absence of chicken manure (or other animal manures), chemical fertilizers may be applied at the rate of 50–100kg/ha 18:46:0 (NPK) or 100–150kg/ha ammonium phosphate (16:20:0, NPK). Immediately after fertilization 3–5cm of water is introduced into the pond. After one week, the same amount of fertilizer is applied and the water level is raised to 10–15cm. The fertilization is repeated after the second week and the water level is raised to 20–25cm. Additional water is added to the pond as necessary to make up for see-page and evaporation losses. Many farmers recommended fertilization every 7 days throughout the fish or shrimp culture period. For further details see Padlan, Ranoemihardjo and Hamami (1975).
    • Yamada (1986) quotes the ‘lab-lab’ fertilization programme of Ballesteros and Mendoza (1976) for milkfish culture in the Philippines: Initial application of 100–200kg/ha 18:46:0 (NPK) onto dry pond bottom, and then allowing water to enter the pond. Additional fertilizer applications of 50–100kg/ha 18:46:0 (NPK) every 10–15 days up to one week before stocking. One week after stocking further application of 15–25kg/ha 18:46:0 (NPK), repeating this dosage every 10–15 days until harvest.
    • Djajadiredja and Poernomo (1973) quote a fertilization programme for the production of ‘kelekap’ (benthic algal complex) for milk-fish ponds in Indonesia: ponds are first drained and the mud thoroughly tilled so as to make it as soft and fine as possible. Fertilizer application of 130kg/ha urea, 65kg/ha triple superphos-phate and 1000kg/ha rice chaff (locally known as ‘sekam’) by scattering evenly over the moist pond bottom. Immediately after fertilizer application water level in the pond raised to 3–10cm. After a good growth of benthic algae is noticed, water level raised to normal level of 30–40cm and fish stocked. Djajadiredja and Natawiria (1965) observed that 8.8–28.1 tons (average 15.6 ton /ha) of ‘kelekap’ could be produced in 2 weeks with an application of 500kg/ha urea. However, chemical fertilizers are rarely used for benthic algal production in Taiwan; preference being rice bran and night soil (Chen, 1973).
    • Proce dure for growing ‘lab-lab’ can also be used for freshwater fish ponds, using an initial fertilization regime of 50kg/ha 16:20:0 (ammonium phosphate, NPK), 22kg/ha 46:0:0 (urea, NPK), and chicken manure at 2000–2500kg/ha (Bautista, 1982; Oandasan, 1982).

21. ‘Lumut’ (benthic grass green flamentous algal complex) for milkfish (C. chanos) / shrimp ponds - General (ASEAN, 1978):

    • Soft mud bottoms with a pH 6.8–7.5 favour growth of ‘lumut’. Pond should first be dried for a 3 day period, and sufficient water allowed to enter the pond to moisten the soil. Moist bottom is then seeded with long filaments from older or existing plants (it usually takes 2–4 weeks from planting until the pond is ready for stocking). After seeding the pond is flooded to 30cm, and 3–7 days later is fertilized with ammonium phosphate (16:20:0, NPK) at a rate of 18–20g/m³ water, applied by broadcasting or by dissolving from a platform placed 10cm below the water level. After one week the water level raised to 40cm, and thereafter weekly applications of fertilizer given at a rate of 9–10g/m³ water until 6 weeks before the crop is harvested. Rows of twigs and small branches should be inserted into the mud bottom (lines 6–15m apart) so as to minimise the destructive dis lodging action of wind and waves on the ‘lumut’. With adequate wind breaks the water can be maintained at a depth of 60cm.

2. Water exchange: For the beneficial effects of liming and chemical fertilizers to be realized in the form of increased phytoplankton production it is essential that the retention time of water in the pond be at least three to four weeks (equivalent to a pond water exchange rate of 5%/day). Water exchange rates greatly in excess of this will result in fertilizer and liming nutrients being flushed out of the pond before they can be used (Boyd and Snow, 1975; Miller, 1976; Boyd, 1986). Excessive water exchange rates may be a major problem in the tropics during the rainy season.

3. Water chemistry: In waters with high calcium concentrations (hard water) and elevated pH, the phosphate applied in fertilizers may be rapidly lost from the water through precipitation as insoluble calcium phosphate, thus rendering it unavailable to the primary autotrophs (Boyd, 1982). It follows therefore that phosphate fertilizer application rates should be higher within hard waters of high pH than in softer water with a more moderate pH (Boyd, 1986). In view of the above relationship, phosphate fertilizers should never be applied at the same time or within one week of liming (Viveen et.al., 1985).

4. Natural soil fertility: Ponds on fertile pasture soils require lower fertilizer application rates than infertile woodland soils (Boyd, 1976). Similarly, rich alluvial soils with a high organic matter content require lower fertilizer application rates than infertile sandy loam soils for the growth of benthic blue-green algae (‘lab-lab’) within brackishwater fish ponds (Tang and Chen, 1967; ASEAN, 1978).

5. Previous pond management: Newly constructed ponds generally require higher initial fertilizer application rates than ponds with a history of fertilization and accumulated bottom sediments (Hickling, 1962; Hepher, 1963; Swingle, 1965; Boyd, 1986).

6. Aquatic weed infestation: Large populations of aquatic macrophytes will compete with phytoplankton for available nutrients and sunlight, resulting in reduced phytoplankton production (Boyd, 1982; Miller, 1976; Boyd, 1986). Weed infestation may be controlled through liming, mechanical cropping, or through the use of herbivorous fish species such as grass carp (C. idella), Tilapia (T. rendalli, niloticus, mossambicus, zillii) or rabbit fish (Siganus sp.).

7. Algal taxonomic composition: Although chemical fertilization stimulates algal productivity, the algal taxonomic composition is generally unpredictable (Boyd, 1986). Recommended dissolved nutrient concentrations favouring predominance and growth of specific algal groups include: diatoms - 20–30:1, N:P (ASEAN, 1978); 10–20:1, N:P (Clifford, 1985); phytoflagellates - 1:1, N:P (ASEAN, 1978), phytoalgal

   plankton (general)	- 4:4:1, N:P:K (Hora and Pillay, 1962)
   	- 4:1, N:P (Swingle and Smith, 1939; Nailon, 1985)
   	- 42:75:1, C:N:P (Hepher and pruginin, 1981)
   	- 50:10:1, C:N:P (Biomass composition; Edwards, 1982)
   Bacteria (general)	- 100:5:1, C:N:P (Growth medium, Edwards, 1982)

8. Fertilizer solubility: A fertilizer will only be effective if it is soluble. Although this is not generally a problem for nitrogen based fertilizers (the majority being very soluble), phosphate fertilizers vary in solubility depending on their particle size and chemical composition (Table 16; Miller, 1976; Boyd, 1979; Hepher and pruginin, 1981). In this respect, liquid fertilizers (if available) are recommended over granular and powdered fertilizers due to their faster solubilization and more uniform distribution of nutrients in the water column (Musig and Boyd, 1980; Davidson and Boyd, 1981).

   Table 16. Percentage dissolution of phosphorus and nitrogen from selected fertilizers after settling through a 2-metre water column at 29°C ^{1 2}

   Fertilizers	Nutrient solubility (%)
   	Phosphorus	Nitrogen
   Superphosphate	4.6	-
   Triple superphosphate	5.1	-
   Monoammonium phosphate	7.1	5.1
   Diammonium phosphate	16.8	11.7
   Sodium nitrate	-	61.7
   Ammonium sulphate	-	85.9
   Ammonium nitrate	-	98.8
   Calcium nitrate	-	98.7

   ¹ Source: Boyd (1982)
   ² The above solubilities are specific for the study in question:solubility also varying with fertilizer particle size and water quality

9. Fertilizer application method and frequency of application: The fertilizer application method used can have a profound effect on the success of a pond fertilization regime. This is particularly true for granular and powdered phosphate fertilizers, which if allowed to come into direct content with the pond bottom will become rapidly adsorbed by the soil particles and so rendering the phosphate unavailable to the planktonic algae. To overcome this difficulty, phosphate fertilizers should be either dissolved in water prior to distribution or applied within floating perforated cannisters, suspended perforated sacks or by placing onto underwater platforms (Figure 13). The latter application methods rely on the gradual dissolution and distribution of the fertilizer by wave action and water circulation within the pond; it follows therefore that such devices should not be placed near the pond outlet (Van der Lingen, 1967; Vincke, 1970; Boyd and Snow, 1975; Davidson and Boyd, 1981; Viveen et.al., 1985; Boyd, 1982; Sanchez and Quevedo, 1987). However, it should be emphasised that the fertilization of brackishwater ponds for the production of benthic algae (ie. milkfish ponds) is radically different from that of freshwater ponds where the main aim is to produce planktonic algae (Chen, 1973; Djajadiredja and Poernomo, 1973; ASEAN, 1978). For the preparation of ponds for benthic production fertilizers are applied directly onto the exposed and dried pond bottom (Table 15).

For the maintenance of the pond primary productivity, fertilizers should be applied on a ‘little and often’ basis, preferably at weekly or biweekly intervals throughout the culture cycle; the residual effect of an applied fertilizer dosage lasting for only two to four weeks depending the water management strategy employed (Hepher, 1963; Miller, 1976; Boyd and Snow, 1975; Hepher and Pruginin, 1981; Viveen et.al., 1985; Vincke, 1985; Boyd, 1982).

a) Underwater platform 1
b) Perforated floating can or basket
c) Suspended perforated sack
Figure 13. Mechanical fertilizer application methods

¹ The base of the platform should be 15–20cm below the water surface, andlocated near the pond water inlet or at the end of the pond from which theprevailing wind comes. A single platform is sufficient for ponds up to 7hawhen plankton is grown. Suggested platform top sizes for ponds of differentsizes include:

3.2.4 Organic fertilization of aquaculture ponds

Organic fertilizers are applied mainly to stimulate the heterotrophic food chain of aquaculture ponds. Although virtually all biological materials can be considered as potential organic fertilizers, the commonest fertilizer used in aquaculture is animal or farmyard manure (ie. farm animal faeces, with or without urine and bedding material). Apart from being a readily available and inexpensive commodity, animal excreta represents a nutrient packed resource containing 72–79% of the nitrogen and 61–87% of the phosphorus originally fed to the animal (Taiganides, 1978). The average nutrient composition of animal manures and other commonly used organic fertilizers has been presented previously (Section 3.13; Tacon, 1987a). However, it must be emphasised at the outset that the nutrient composition of animal manure is highly variable (depending on the diet of the animal, the age and species of the animal, the type and proportion of bedding material present, and the handling and treatment of the manure prior to usage), and consequently each manure source must be considered as being unique and chemically analysed accordingly. Sadly, the majority of published aquaculture production trials involving the use of animal manures rarely report nutrient analyses of the ‘pig’, ‘poultry’ or ‘cattle’ manure used, the presence or not of bedding material, or whether the quantities of manure applied to the pond were on a dry or fresh weight basis.

3.2.4.1 Effect on pond productivity and fish/shrimp production

In contrast to chemical fertilizers which act directly on the autotrophic food chain, organic fertilizers act mainly through the hetero-trophic food chain by supplying organic matter and detritus to the pond ecosystem; the manure serving principally as a substrate for the growth of bacteria and protozoa, which in turn serve as a protein rich food for other pond animals, including the cultured fish or shrimp (Figure 14). Whereas autotrophic production within fertilized ponds is limited by available solar energy (Table 17), heterotrophic production will depend upon the carbon and nitrogen content of the added manure and its consequent susceptibility to microbial decomposition (Schroeder, 1978, 1980; Wohlfarth and Schroeder, 1979). The C:N ratio of the applied manure will determine its rate of bacterial decomposition in water and hence the time lag between application and increased heterotrophic pond productivity; manures with a low C:N ratio (< 50; animal manures, green weeds, grass, oilseed meals) being more rapidly decomposed by bacteria than wastes with a high C:N ratio (> 100: straw, sugar cane bagass, sawdust; Tacon, 1987a, Sturmer, 1987). Schroeder (1980) suggests that the ideal C:N ratio for a bacterial growth medium is about 20:1. It follows from the above that the smaller the particles of organic matter the faster will be the colonization and decomposition by bacteria and protozoans (Geiger, 1983); for example, fresh animal manure readily disintegrates in water into colloidal particles. Schroeder (1980) estimates that the aerobic digestion of organic matter by bacteria fixes about 20–50% of the substrate carbon into new bacterial biomass; the yield of bacterial biomass obtained by aerobic digestion being about 10 times higher than by anaerobic digestion (McCarty, 1972). According to Cassinelli et.al., (1979) for each gram of organic matter decomposed, 1.2g of oxygen is consumed, and that for each gram of carbon fixed during photosynthesis, 2.6g of oxygen is produced. These authors concluded that the major source of oxygen to a shrimp pond was derived through algal photosynthesis and that the major oxygen sink was algal and bacterial respiration (cited by Pruder, 1986).

Figure 14. Fate of applied organic fertilizer in aquatic systems (adapted from Edwards, 1982; Delmendo, 1980 and Moore, 1986)

Table 17. Primary productivity and fish yields attainable within chemically fertilized and manured ponds in Israel ¹

¹ For standing water ponds receiving no supplemental feeds (Schroeder, 1980)
² Ammonium sulphate and superphosphate applied once every 2–3 weeks at60kg/ha
³ Manure application 6 days/week, at a daily dry organic matter loadingrate equivalent to about 3% of the fish biomass (field dry chicken manureat a rate of 100kg organic matter/ha/day).

The beneficial effects of organic fertilization on natural pond productivity are well illustrated by the studies of Schroeder (1980) and Rappaport, Sarig and Bejerano (1977), and their results are summarised in Table 18. For additional information on the stimulatory effect of manure on pond biota productivity see Tang (1970), Noriega-Curtis (1979), Olah et. al., (1986), ASEAN (1978), Malecha et. al., (1981), Lee and Shleser (1984), Barash and Schroeder (1984), Wyban et. al., (1987), Garson, Pretto and Rouse (1986), and Zhang, Zhu and Zhou (1987).

Intense organic and chemical fertilization of aquaculture ponds has resulted in fish and shrimp yields as high as 5–10 tons/ha/year or 15–32 kg/ha/day with no supplementary feeding (Fish- Tang, 1970; Schroeder, 1974, 1980; Schroeder and Hepher, 1979; Moav et. al., 1977; Wohlfarth, 1978; Buck, Baur and Rose, 1978; Delmendo, 1980; Nash and Brown, 1980; Edwards, 1980; Maramba, 1978; Djajadiredja and Jangkaru, 1978; ADCP, 1979; FAO, 1983; Vincke, 1985; Zweig, 1985; Plavnik, Barash and Schroeder, 1983; Behrends et. al., 1983; Shrimp- Wyban et. al., 1987, Lee and Shleser, 1984). However, these high production levels can only be achieved by using appropriate management controls, and paying particular attention to fish/shrimp stocking density and species selection (Schroeder, 1978; Wyban et. al., 1987). For example, Schroeder (1978) correlated fish yields in ponds receiving only cattle manure and chemical fertilizers with stocking density and found a linear relationship up to 9300 fish/ha (ie. the carrying capacity of the pond; Figure 15). For each fish stocked, up to 9300 fish/ha, an annual yield of 0.75kg fish was obtained (as compared with an annual yield of 1kg for fish ponds employing conventional pelleted feeds; Hepher and Schroeder, 1974). These results also indicated an efficient manure conversion efficiency into new fish tissue; for every kg of fish produced, approximately 3–3.5kg of manure dry matter was used (conversion efficiency cited by Hepher and Pruginin, 1981). Wohlfarth and Schroeder (1979) report a conversion efficiency of 2.7 and 3.5 for cattle and chicken manure for manuring trials conducted at Dor, Israel, with a polyculture of common carp, silver carp, tilapia and grass carp. By contrast, Garson, Pretto and Rouse (1986) report a shrimp (P. vannamei/ P.stylirostris) conversion efficiency of 17 and 20 for chicken and cow manure, respectively (conversion efficiencies calculated on a manure dry weight basis and for whole shrimp). Recalculation of the data of Wyban et. al., (1987) with shrimp (P. vannamei) ponds receiving only feedlot cattle manure shows a conversion efficiency (dry manure:whole shrimp) of 21 and 11 for shrimp at stocking densities of 5/m² and 15/m², respectively; these authors also reported that the carrying capacity of their manured ponds receiving 1800 kg feedlot manure/ha/week was equivalent to about 1700 kg shrimp/ha.

Table 18. (a) Standing crops of phytoplankton, zooplankton, chironomid worms and bacteria in manured and non-manured ponds with or without fish in Israel ¹

¹ Source: Schroeder (1980) - water temperature 9–15°C, using cowshedmanure and a common carp, tilapia and silver carp polyculture
² Phytoplankton retained on a 50 micron net, grams dry weight/m^3
3 Zooplankton retained on a 150 micron net, grams dry weight/m^3
4 Bacteria concentration within the pond water column (for pond bottomswith an organic matter content greater than 1%, the bacterialconcentration is 100–1000 times higher on the pond bottom than in theoverlying water column; Schroeder, 1978).

(b) Natural food organisms found in water and bottom soil of manured and non-manured fish ponds in Israel ¹

¹ Source: Rappaport, Sarig and Bejerano (1977)
² Dried manure allowed to stand covered with water for 7 days, and appliedat a rate of 5kg dry matter/ha/day
³ Dung and excreta containing about 10% dry matter, application as forchicken droppings
⁴ Fresh cow dung, also containing remnants of feed and coarse bedding,treated as for poultry droppings
⁵ 20kg ammonium sulphate and 15kg superphosphate/ha/week.

Figure 15. Relationship between polyculture stocking density and fish yield in standing water earthen ponds receiving fertilizer inputs only (Schroeder, 1980)

If maximum benefit is to be gained from the wide variety of live food organisms available within a well fertilized pond (ie. phytoplankton, zooplankton, bacterial enriched detritus, macrophytes, benthic algae and animals) it is essential that these ponds be stocked with fish and/or shrimp with diverse feeding habits (Stickney, 1978; Schroeder, 1980; Wohlfarth and Schroeder, 1979; FAO, 1983; Vincke, 1985; Malecha et. al., 1981; Zweig, 1985). Fish polyculture strategies date back to the Chinese Tang Dynasty (7th centuary A.D; Zweig, 1985) and in China rest on three basic principles (FAO, 1983):

• “complete use of the pond, both in depth, from the surface to the benthic zone and over its entire surface area;

• complete use of all types of natural food present in the pond: phyto- and zooplankton, benthos, aufwuchs, detritus, aquatic plants, and

• taking advantage of mutual benefits while avoiding competition for food.¹ Several different species are therefore reared together in the fattening pond. Depending on the type of food available locally, one or two of the main species of Chinese carps are chosen: silver, bighead, grass, black or mud. They are then combined with complementary secondary species on the basis of the principles set out above and the ecological requirements of the species considered, for example:

  1. since the droppings of the grass carp are rich in undigested plant fibres, they help the development of plankton which feeds silver and bighead carps ;
  2. to control molluscs, 75–100 black carps/ha are added to the pond, while, to control small fish and red shrimp, 450–600 carnivorous fish may be added, if the pond is drained annually;
  3. the common carp scours the bottom of the pond to obtain its nourishment and this helps aerate the sediment, oxidize organic matter, recycle minerals and finally encourages the development of plankton and the growth of plankton-feeding species, and
  4. nevertheless, competition may develop between the common carp and the mud carp, the silver and bighead carps, or the silver and mud carps, which makes it necessary to limit the number of one or the other of these species (common carp: 150–225kg/ha; silver carp: 300–450 kg/ha)”.

Out of a total of 25 fish species cultured in China, nine species have sufficiently different feeding habits that they can be cultured together at the same time in a single pond: grass carp (C. idella) and wuchang fish (M. amblyocephala) feed on terrestrial plants and aquatic macrophytes; silver carp (H. molitrix) and bighead carp (A. nobilis) feed mainly on phytoplankton and zooplankton, respectively; black carp (Mylopharyngodon piceus) feed on molluscs (snails); mudcarp (Cirrhinus molitrella) feed on bottom detritus; and common carp (C. carpio) feed on benthic invertebrates and most of the above food items with the exception of plankton (Zweig, 1985). Table 19 shows the natural feeding habits of adult tilapias and other important fish and prawn species. Species ratios which have been found to give satisfactory results in fertilized ponds include:

1. Common carp:tilapia (aureus): silver carp, 5:2.5:1.5; total of 4500–9500 fish/ha (Israel - Schroeder, 1978, 1980).
2. Silver carp:bighead carp:grass carp:common carp, 65:1:4:12, with a combined density of about 5500/ha, together with freshwater prawn (M. rosenbergii) at density of 7.9/m² (USA - Malecha et. al., 1981; for other prawn polyculture studies see Wohlfarth et. al., 1985, Rouse, Naggar and Mulla, 1987, and Cohen, Ra'anan and Barnes, 1983).
3. Silver carp:bighead carp:grass carp:wuchang fish:crucian carp(Carassius carassius):common carp, 4500:1500:4500:3000:3000:1500/ha, total 18000 fish/ha (China - Shan et. al., 1985).
4. Common carp (50–70%), silver carp (20–30%), bighead carp (10%), grass carp (5–10%), and sheat fish (Silurus glanis; Hungary, ADCP, 1984).
5. Silver carp:bighead carp:grass carp:common carp, 7500:1550:4500:1500/ha, total 15000 fish/ha; silver carp:wuchang fish;crucian carp:bighead carp: grass carp:common carp, 4500:3000:3000:1550:4500:1500/ha, total 18000 fish/ha (China - Zhang, Zhu and Zhou, 1987; for other Chinese polyculture ratios see FAO, 1983).
6. Silver carp:bighead carp:grass carp:tilapia (niloticus, males): tilapia (aureus, males), 2500:250:150:7500:5000/ha, total 15400 fish/ha (Alabama USA - Behrends et. al., 1983).

¹ For example, in Israel Yashouv (1971) reported a common carp production of 390kg/ha in monoculture and 714kg/ha in polyculture with silver carp (silver carp production 1923kg/ha; both ponds receiving equal inputs of inorganic fertilizers and poultry manure). Yashouv explains the improved growth to be the result of a “positive (synergistic) interaction on the basis of increased food sources. Each of the fish species processes a food source; thus making it available to the other. The faecal pellets of silver carp, which are rich in partially digested phytoplankton, make this food source available to common carp which otherwise could not utilise the phytoplankton. The common carp, by digging and ploughing the pond bottom, release into the water minute organic matter, which is then strained out and utilised by silver carp”.

The ultimate choice of species ratio and stocking size will depend upon the type of farming activity envisaged (rural/subsistence or commercially oriented farming activity), the availability and cost of fertilizers and feeds, and on the natural productivity of the water body in question. For information on the calculation of fish polyculture ratios and stocking densities see FAO (1983) and Horvath, Tamas and Tolg (1984).

Table 19. Natural feeding habits of some pond cultured fish and prawn species

¹ Bowen (1982)
² Schroeder (1980)
³ King and Garling (1986)
⁴ Malecha et. al., (1981)

3.2.4.2 Manure fertilization through straight manual application

The stimulatory effect of an animal manure on natural pond productivity will be determined to a large extent by its method of distribution and application (ie. quantity and frequency of application) to the pond. The better the distribution of the manure over the pond area the better the fertilization effect achieved (Woynarovich, 1985; Delmendo, 1980; Edwards, 1982). Furthermore, manures which produce fine colloidal particles are more rapidly colonised and decomposed by bacteria, and consequently will be more effective, than manures presented in large lumps or heaps (Hepher and Pruginin, 1981). Woynarovich (1979) found that when soft fresh animal manure was mixed with pond water and repeatedly spread over the entire pond area that sufficient amounts of carbon compounds were liberated to maintain a high primary productivity. This was believed to be due to the fact that approximately 30% of the total dry matter content of the liquid cow manure existed in a colloidal state and thus acted as an ideal substrate for bacterial and protozoan growth on the pond bottom and within the water column (Moav et. al., 1977). Similarly, Schroeder (1980) reported that as much as 40% of the total solids of fresh cow manure remained in suspension in the water column; 50–60% of which being in the form of inorganic materials. However, he also noted that approximately 90% of the coarse organic matter settled to the pond bottom after one to two hours, and that sediment accumulations of more than a few mm resulted in the development of anaerobic sediment conditions. It follows from the above that there is a maximum amount of manure that a pond can aerobically digest/unit area/unit time; the addition of manure above this maximum level leading to the accumulation of organic matter on the pond bottom and the development of undesirable anaerobic interstitial conditions (Edwards, 1982). According to Schroeder (1980) the maximum amount of manure that a pond can safely digest without undesirable anaerobic effects is about 100–200kg manure dry weight/ha/day or 70–140kg organic matter/ha/day (for Israeli pond conditions). These values correspond approximately to the manure produced from 100–200 pigs weighing 100kg each/ha/day, 15–30 cows weighing 500kg each/ha/day, or 2000–4000 poultry each weighing 2kg each/ha/day (Edwards, 1982). To obviate the possible dangers of water deoxygenation within manure loaded and eutrophic ponds (due to unchecked peaks in bacterial growth and phytoplankton blooms), manures should be added as frequently as possible, at least daily, on a little and often basis (Hepher and Pruginin, 1981; Wohlfarth and Schroeder, 1979; Schroeder, 1978; Woynarovich, 1979). Although the oxygen demand of the manure itself is not great if the manure is evenly distributed over the pond surface, it is recommended to apply manure to a pond during mid-morning when oxygen levels are rising rapidly due to photosynthesis; this in turn would minimise the oxygen demand caused by the bacterial breakdown of the manure itself during the critical pre-dawn hours (Woynarovich, 1980; Edwards, 1982). In addition, since the manure requirement of a pond will depend upon the dietary live food requirements of the fish/shrimp biomass present, it follows that the manuring rate will have to be increased (up to a maximum safe level) with increasing fish biomass or standing crop (Hepher and Pruginin, 1981). Figure 16 shows the relationship between total standing crop and daily manure requirement obtained by Wohlfarth (1978) for Israeli fish ponds. Examples of manure fertilization programmes which have been employed by other workers are shown in Table 20. It must be remembered, however, that the manuring rates shown are pond and farm specific, and as such should only be used as tentative ‘guidelines’ by persons wishing to develop their own pond fertilization programmes.

Figure 16. Relationship between manure requirement and standing crop in Israeli fish ponds (Wohlfarth, 1978)

Three basic methods are currently employed for the distribution of animal manure to fish or shrimp ponds (Woynarovich, 1979):

• The dilution of the manure on land and the distribution carried out by hand from the shore or from a small boat. This method is normally used for small ponds (Figure 17, A1–3).

• Soft manure is shovelled into a basket of parallel iron rods (approx. 2–2.5cm apart), suspended 10–20cm below the water line, attached to the side of a boat, and dispersed as the boat moves and forces water into the basket (Figure 17, B1).

• The use of a pump built into the bottom of a boat; the manure is shovelled into a hopper, diluted with pumped water and sprayed out into the pond through a flexible hose (Figure 17, C1).

Figure 17. Organic manure distribution methods (Woynarovich, 1985)

Table 20. Examples of manure fertilization programmes for pond fish and shrimp

FRESHWATER