Larvae can either be reared extensively in fertilized pond or under intensive conditions in tanks.

Extensive larval rearing

Ponds are often used for extensive larval rearing. Production costs are lower and larvae have a faster growth than larvae reared intensively in tanks possibly owing to better nutrition resulting from a more varied diet and to greater natural prey availability throughout the day (Russel and Rimmer, 2004).

For extensive rearing, ponds must be drained and dried until the bottom soil cracks to release toxic gases, oxidize mineralized nutrients and eradicate some pests and predator (Kungvankij et al., 1985). The pond bottom should then be disinfected and neutralized with agricultural lime at 2 mt/ha, where after the pond can be filled with water filtered through a 300–500 µm mesh to avoid the introduction of predators (Rutledge and Rimmer, 1991), and fertilizers can subsequently be applied. A combination of inorganic and organic fertilizers is recommended. The most commonly used fertilizers in Australia are diammonium phosphate (DAP) and lucerne pellets (Russel and Rimmer, 2004). The fertilization schedule adapted from Rimmer (1991), presented in Table 3, can be used as an example. Dissolved oxygen, salinity, pH, temperature and zooplankton density should be monitored daily and fertilization rates should be adjusted accordingly.

Two days after hatching larvae are stocked at 400 000 to 900 000 fish/ha into the pond, to coincide with the peak densities of smaller zooplankton (rotifers and copepod nauplii of 50–100 µm) (Rutledge and Rimmer, 1991). This peak may occur between one and two weeks after pond filling depending on the location. It is therefore important to monitor the pond. If stocking occurs too late, large zooplankton may feed on fish larvae.

As the larvae grow, they feed on larger zooplankton such as adult copepods and cladocerans. As zooplankton densities decrease as a result of predation, the fish switch to benthic food sources, principally midge larvae “blood worms” (Rutledge and Rimmer, 1991).

When fish reach 20 mm or more (about two weeks after stocking), they can be harvested and transferred to the nursery facility.

Early weaning onto commercial microdiets has been investigated to reduce the dependence on live food and to reduce production costs. In a recent study, 2 DPH larvae were fed directly on a microdiet without rotifers but the final weight and the survival rate at 28 DPH were low (1.3 mg and 0.1 percent respectively) (Curnow et al., 2006). Low digestive enzyme activity at this stage may explain the low survival and growth rates. Replacement of live feed must therefore be gradual and take place over several days. Weaning protocols using inert and live diets (co-feeding) allow for earlier and more efficient weaning with higher growth and survival than feeding either live feeds or microdiets alone. It has been observed that digestion of artificial feed occurred before metamorphosis when the fish were fed in conjunction with rotifers because the latter generate sufficient proteolytic activity in the intestine (Southgate and Lee, 1993). Weaning on to dry feeds is usually accomplished before 25 DPH (around 11 mm TL) (Tucker, 2005). However, Barlow, Williams and Rimmer (1996) reported better survival if weaning is delayed until the fish are at least 15 to 20 mm TL. Recent advances in microdiet formulation allows the fish to be weaned earlier and Curnow et al. (2006) suggested that co-feeding barramundi larvae on high quality microdiets should start when the larvae reach 5 mm (around 15 DPH. An example of a feeding schedule with an early co-feeding protocol is given in Table 5b. Appropriate particle size of microdiets is essential (Walford et al., 1991) for good survival and growth.