(Coche 1978, 1979). The utilization of a high initial biomass then becorries
not only possible but also necessary if production costs are to be minimized
and net profits maximized.
Researchers have used either commercial fish feeds or self-made com-
pounded feeds. Commercial fish feeds have a very high protein level (e.g.,
Purina Trout Chow, 40%; Trouvit, 46%), a balanced composition including
essential minerals and vitamins (at least for the target fish species), and a
high price. Self-made feeds, however, have used mixtures of locally available
ingredients, giving lower protein levels (usually 20 to 30%) and lower costs.
Several empirical feed formulations have been evaluated but none has been
developed from scientific data.
Table 9 gives examples of artisanal feeds which have been found satis-
factory under various local conditions. Their composition is largely based on
relatively cheap materials, such as rice bran and cotton seed oil cake, enriched
with either fish meal or blood meal. The actual price of the mixed ingredients
is now about US$0.15 to 0.20/kg, and the cost of pelleting probably adds
another US$O.lO/kg. The Campbell formula B4 (Table 9) is particularly
attractive because of its relatively high protein content (25%), low cost
(currently US$O.2O/kg) and high efficiency (FCR = 2.0), at least for S.
niloticus.
In the Philippines, Pantastico and Baldia (1979) have used a feed com-
bining 20% ipil-ipil leaf meal (Leucaena leucocephala, Leguminoseae),
20% fish meal and 60% rice bran for S. mossambicus cage culture. This
feed contained 27.3% dry weight of protein and cost only US$O.O9/kg.
Unfortunately the adverse conditions under which these experiments were
conducted (small-size mesh, low density, low biomass, high feed ration)
prevented a realistic demonstration of the efficiency of this feed (FCR =
4.0).
For S. niloticus in particular, an omnivore with a vegetarian tendency
in its natural environment, Coche (1977) stressed that the ideal artificial
diet a priori should contain a relatively high percentage of carbohydrates
as an energy source. This appears to have been confirmed by Campbell
(197813) for S. niloticus in Lake Kossou, Ivory Coast. In his B4 formulation
(Table 9) the 25% protein includes at least some of animal origin (i.e.,
4% fish meal). The addition of a mineral/vitamin premix to B4 did not
improve the results.
The feeding rate is usually quantified as the daily feed ration (DFR),
the amount of feed (wet or dry) being fed daily (generally six days a week)
to the fish, expressed as a percentage of the best available estimate of the
fish biomass (%B). Such estimates, based on cage sampling, are usually made
every 15 to 30 days.
It is well known in fish husbandry that food requirements per unit
weight of fish decrease as the fish increase in size, but for the tilapias there is
little information on this. In intensive cage culture of S. niloticus in the
Ivory Coast, it was shown that the DFR had to be decreased from 6 to