POLYCULTURE OF SILVER BARB, PUNTIUS-GONIONOTUS (BLEEKER), NILE TILAPIA, OREOCHROMIS-NILOTICUS (L.), AND COMMON CARP, CYPRINUS-CARPIO L., INVIETNAMESE RICEFIELDS - FEEDING ECOLOGY AND IMPACT ON RICE AND RICEFIELD ENVIRONMENT

Rice production, ricefield environment and the feeding ecology of fish were studied in an experiment conducted at a rice-fish station in the Mekong Delta, Vietnam. In total, six treatments (three replicates) we re investigated: four different polyculture combinations of small size d silver barb, Puntius gonionotus (Bleeker, Nile tilapia, Oreochromis niloticus (L.), and common carp, Cyprinus carpio L.; one treatment wit h pre-grown fingerlings; and a control treatment in which no fish were stocked. No insecticides or fungicides were utilized before or during the experiment. Frequent fertilization of the water and a low rice pl ant biomass during the early vegetative growth phase stimulated the de velopment of phytoplankton and zooplankton. The total weed biomass was low (maximum = 5.3 g dry weight m(-2)) and not significantly (P < 0.0 5) different between the treatments, A major component of the silver b arb diet consisted of rice plants and accessible grains. However, the introduction of silver barb only had a significant effect on the numbe r of rice tillers in the ratoon crop and not on the paddy yield. The q uantitative differences in the diets of tilapia and common carp were m inimal: both species fed mostly on detritus. Ricefields without silver barb produced the highest paddy yield (3120 kg ha(-1)). The total yie ld of introduced fish increased after increasing the stocking density of silver barb from 319.9 to 494.1 kg ha(-1). The highest fish yield ( 541.8 kg ha(-1)) was obtained by stocking pre-grown fingerlings in the ricefields.