Results
The descriptive statistics (Table 2) showa great di¡erence
between the minimum and maximum C. carpio
net production. This can be directly related with the
huge range for the variables ‘SGR’and ‘survival rate’
(Tables 3 and 4), as well as ‘C. carpio mean daily gain’
and ‘C. carpio harvesting weight’ (Table 2).
The SGR multiple regression solution (Table 5) explains
nearly 66% of the observed variance in the
SGR variable (variables that were not signi¢cant
(Po0.05) were not included in the table). The main
variable a¡ecting SGR C. carpio is the C. carpio initial
weight: SGR decreases with increasing C. carpio initial
weight. Furthermore, at the beginning of the rice
crop (the vegetative phase), when the rice stands less
dense andwhen rice biomass is still low, SGRC. carpio
values are higher. Later onwhenthe rice has a denser
stand and higher biomass (reproductive phase), the
SGR is lower. A high O. niloticus SB decreases SGR
C. carpio, suggesting interspeci¢c competition.
The multiple regression for the survival rate of C.
carpio (Table 6) explains about 70% of the observed
variance (variables that were not signi¢cant
(Po0.05) were not included in the table). The results
are not surprising. Increased C. carpio stocking density
and increased wild ¢sh biomass have a negative
impact on C. carpio survival rate. Higher C. carpio
stocking weight had a positive impact on the ¢sh’ survival
rate. Survival rate also depended on the season.