The results from eight rice^¢sh experiments (Table1)
carried out at the rice^¢sh research station at the Co
Do state farm, Can Tho province,Vietnam, between
November 1995 and October 1999, were used to
analyse the SGR and survival rate of C. carpio in intensively
cultured rice ¢elds characterized by two
short-duration, highyielding direct-seeded rice crops
and intensive fertilizer use.
The Co Do area is characterized by about 2500
hours of sunshine a year and an annual rainfall of
approximately 1600mm. There are two main seasons:
a dry season from December to mid-May and a
wet season from mid-May to November. The region is
also characterized by yearly £ooding (0.5^1-m depth)
from August to November, during which rice cropping
is impossible and ¢sh culture risky. Farming
in the Co Do area is characterized by an intensive
direct-seeded double rice cropping system: a dry
season crop (December^March) followed by a wet
season crop (May^August).
All plots used in the experiments measure about
650 m2 and have a rice ¢eld area (550m2) and a 1-m
deep L-shaped trench (100m2) that serves as ¢sh refuge.
All plots are surrounded by a dike, and have
their own water inlet and outlet. A detailed description
of the study area, experimental station, rice^¢sh
¢eld layout and soil andwater resources can be found
in Rothuis (1998).
Half of the experiments used for the study took
place in the wet season; the other half took place in
the dry season (Table 1). Experiments in the wet season
included a rice crop followed bya rice ratoon crop
(ratooning is the ability of rice plants to regenerate
new tillers after harvest; Chauhan,Vergara & Lopez
1985). The experiments in the dry season included a
rice crop ^ followed in only one experiment (experiment
7) by a period during which ¢sh remained in
the trench without any access to the ¢eld. In the ¢rst
four experiments, the rice variety IR56279 was used,
in the other IR62032. Both are short-duration varieties;
they can be harvested at about 100 days after
seeding (DAS).Whereas we applied inorganic fertilizer
to the rice at normal rates (141kg per rice crop
per ha urea,141kg per rice crop per ha di-ammonium
phosphate and 50 kg per rice crop per ha potassium
chloride), we did not apply organic manure. All ¢elds
were hand-weeded and, when necessary, insecticides
or fungicides were applied.
Thewater level on the rice ¢eldwasmaintained between
10 and 15 cm, except in the ratoon crop when
water levels were between 30 and 40 cm. B. gonionotus
(Bleeker), C. carpio L. and O. niloticus (L.) were
stocked in all eight experiments. Trichogaster pectoralis
Regan was stocked in three experiments, whereas
Osphronemus goramy Lacepe'dewas stocked in one experiment.
All ¢sh were stocked at about 20 DAS. A
fewdays before rice seeding, trenches of all plotswere
treated with15 ppm derris root (5% rotenone) to eradicate
wild ¢sh. A 5-mm screen at the ¢eld water
inlets and a 10-mm screen at the main water gate of
the station helped to prevent other ¢sh from entering
the ¢elds.