articles
eggs were then incubated on different
test substrate materials. A bird’s feather
was used to spread the eggs evenly on
the substrates in bunches of 600-1 200
eggs. The natural substrates tested were
roots of Nile cabbage (Pistia stratiotes),
water hyacinth (Eichhornia crassipes), pond
weed (Ceratophyllum dermasum) and green
grass (Commelina sp.) leaves. The artificial
substrates tested were kakaban mats,
sisal mats, papyrus mats and nylon mats,
all with an equal surface area of
1 350 cm2. Concrete slabs were used
as the control as they are widely used
as a hatching medium in Kenya. All the
test substrates were put in flow-through
concrete troughs. Water temperature
was maintained at 22.9 ±1.1 °C, D.O.
at 5.9 ± 0.2 mgl-1, pH at 7.0 ± 0.5 and
conductivity at 648 ± 10.2 μs throughout
the period of the experiment. The
percentage of hatching was obtained by
using the formula as stipulated in Viveen
et al. (1985). The cost of using each of
the hatching substrates was assessed by
accounting for all expenses incurred for
each of the methods assessed. Completely
Randomized Block Design (CRBD) was
used to allocate 600-1 200 eggs into each
of the experimental units.
The resulting hatching rates were
analyzed by two-way analysis of variance
(ANOVA). Multiple comparison analysis
was used to assess any heterogeneity.
The Multiple Range Test was used to
discriminate among means (Zar 1984).
The tests were at the 95% significance
level (p<0.05). The two-way ANOVA was
also carried out to assess for differences
between the hatching rates of eggs from
different female spawners.
Results
The hatching rates of the eggs on natural
substrates were significantly higher than
on the artificial ones. The mean rates were
66.2, 54.0, 49.7 and 13.0% for Pistia, green
grass leaves (Commelina sp.), E. crassipes
and C. dermasum, respectively (Figure 1).
The difference between the means of
the four data samples was statistically
significant at the 95% confidence level.
Substrates
Figure 1. The percentage hatching rates using different test substrates.
The multiple range tests revealed that
all the means were significantly different
from each other except for one pair - the
difference between green grass leaves and
water hyacinth roots was insignificant.
Three groups of means were identified,
with C. dermasum performing between
15-18%, the green grass leaves (Commelina
sp.) and E. crassipes roots rates falling
between 43-60%, and the Pistia rates
ranking the highest at between 62-76%.
The difference between the means of
the hatching rates of all the artificial test
substrates was also statistically significant.
The hatching rates of the eggs on this
category of substrates were generally low,
with the highest being 18.6% for both the
concrete slabs (control) and sisal mats.
The other mean rates were 4.0, 11.8, and
12.2 for nylon, kakaban, and papyrus mats,
respectively (Figure1). The difference
between the means of the four data
samples was also statistically significant
at the 95% confidence level.
Multiple range tests revealed that all the
means for the artificial substrates were
statistically different from each other.
However, for two pairs of means - the
concrete slab and sisal mats; and kakaban
and papyrus mats - the difference within
each pair was insignificant. Three groups
of means were identified in ranking the
substrates performance. Nylon performed
below 8% and was ranked the lowest,
kakaban and papyrus mats performed
between 8-16%, while sisal mats and
concrete slab performed between 15-25%
and were ranked in the highest group.
There was a significant cost difference
between the use of the artificial
substrates and the natural substrates.
The cost of using the artificial substrates
was higher than for the natural ones.
The cost ranged between US$ 0.12-0.27
for artificial substrates and US$ 0.01-0.03
for natural substrates (Table 1). Direct
costs accounted for much of the cost
incurred in the use of artificial materials,
in contrast to the indirect costs incurred
in the use of the natural materials. The
indirect cost was the opportunity cost
of the man-hours spent in collecting the
substrate materials.
Discussion
The results indicate that the natural
substrates performed better than the
artificial substrates. This was probably
24 NAGA, WorldFish Center Quarterly Vol. 28 No. 3 & 4 Jul-Dec 2005