The fish production parameters of five polyculture combinations, consisting of small and large silver barb, Puntius gonionotus (Bleeker), small and large Nile tilapia, Oreochromis niloticus (L.), and small common carp, Cyprinus carpio L., fingerlings in three replicates, were investigated in a rice‐fish culture experiment (duration 149 days) conducted in the Mekong Delta, Vietnam. The survival rate was not significantly (P > 0.05) affected by the polyculture combination, but when grouped according to species, the mean survival of silver barb and tilapia was 64.3% and 63.7%, respectively, significantly higher than the mean common carp survival rate (33.4%). The growth of silver barb and tilapia was proportionally related to the stocking density, probably because of intraspecific competition and a synergistic interaction between silver barb and tilapia. The growth of common carp was not significantly different among the polyculture combinations. The highest net production (474.1 kg ha−1) was obtained in the polyculture combination consisting of 80% small‐sized silver barb fingerlings, but the fish was not marketable at that time. In concurrent rice‐fish culture, it is recommended to raise large silver barb fingerlings. Small tilapia can be polycul‐tured with silver barb, provided a stocking density lower than 1400 ha−1. Common carp is considered less suitable because of a limited tolerance for the water quality conditions in the ricefield and the large size required by the market.

In the freshwater area of Vietnam's Mekong Delta, Oreochromis niloticus (L.), Barbodes gonionotus (Bleeker) and Cyprinus carpio (L.) are often reared together in rice fields. In this study, we report the results of eight such polyculture experiments, examining variables affecting the specific growth rate and the average daily food energy consumption of individual O. niloticus. The standing biomass of O. niloticus and the wild fish biomass had a negative impact on the specific growth rate, whereas added pig manure, extra feed and inorganic fertilizer had a positive effect. The standing biomass of O. niloticus and C. carpio, and the wild fish biomass had a negative impact on the consumption of natural feed. We inferred that O. niloticus mainly lacks food in rice fields. This results in intraspecific competition. As in rice fields, C. carpio has basically the same feeding niche as O. niloticus; we found interspecific competition between the two species. O. niloticus consumed relatively more food in the vegetative phase of the rice crop, probably because of a higher abundance of phytoplankton. To get higher growth of O. niloticus, we recommend that farmers increase nutrient inputs and stock O. niloticus according to the size of the trench adjoining the rice field and not according to the size of trench plus field as is generally done.

In the freshwater area of Vietnam's Mekong Delta, Barbodes gonionotus (Bleeker), Cyprinus carpio L. and Oreochromis niloticus (L.) are often reared together in rice ¢elds. The survival rate and speci¢c growth rate (SGR) of C. carpio showed a large variance. To understand the underlying reasons for this variation, we used multiple regressions for the variables SGR and survival rate, using the results of eight experiments in such polycultures at the Co Do rice^¢sh experimental station. The SGR of C. carpio declined with increasing rice density. Interspeci¢c competition with O. niloticus also had a negative e¡ect on C. carpio SGR.When fertilizer or manure is added to the trench of the rice^¢sh ¢eld or when extra food can be given to the ¢sh, O. niloticus seems more appropriate than C. carpio. The survival rate of C. carpio depended on season, C. carpio stocking weight and density and the biomass of wild ¢sh species. To obtain higher survival rates, one should stock fewer ¢sh but of a larger size, after thorough eradication of wild ¢sh species.

Trichogaster pectoralis Regan is often cultured successfully in rice fields in South‐east Asia. However, since the advent of high‐yielding rice varieties and intensive rice cropping, several reports suggest poorer growth performance of the species. In this paper, we investigated the growth performance of T. pectoralis in rice–fish systems characterized by intensive rice culture and extensive to semi‐intensive polyculture of Oreochromis niloticus (L.), Cyprinus carpio L. and Barbodes gonionotus (Bleeker). In these systems, T. pectoralis recorded negative net productions. This was mainly the result of a negative specific growth rate (SGR) when fish had to stay in the trench (−0.519% body weight day−1) or when fish had free access to the rice crop (−0.081). Only when harvested rice plants were allowed to generate new tillers (ratooning) did T. pectoralis perform slightly better (0.453% body weight day−1). The SGR of T. pectoralis was low owing to a lack of food in intensive rice systems. The other species, all with SGR values higher than 1% body weight day−1, clearly had a competitive advantage over T. pectoralis. We concluded that intensification of rice cropping did indeed result in poor T. pectoralis growth. Successful T. pectoralis culture is only possible in areas of traditional rice cropping.