Abstract
This study was carried out in Mymensingh District of Bangladesh from June 2002 to February 2005 in three phases with the overall objective of improving the efficiency of nutrient use in integrated pond-dike production system in the region. In the first phase, a Participatory Community Appraisal (PCA) and a State of System (SOS) workshop were conducted to obtain an overview of the community in terms of livelihoods and to identify constraints, opportunities and researchable issues in pond-dike system. The ponds, both in rural and peri-urban areas were found to be essential for people's livelihoods and priority areas for research were identified as status of integration of fish and dike crop, benefits of pond-dike production, selection of appropriate fish for the pond and crops for the pond-dike. In the second phase, both low and high input pond-dike systems as well as rich and poor categories of farms were studied. In low input system, carp polyculture was practised using on-farm waste at an irregular basis, whereas in high input system Thai pangas (catfish) monoculture was practised using mainly pelleted feed. The pond input and output (expressed in Nitrogen) were 2,024.67 kg N ha1 and 690.33 kg N ha1, respectively in the high input system while the same were 73.67 kg N ha1 and 36.0 kg N ha1, respectively in the low input system and the difference was highly significant (P < 0.001). Different types of vegetable were grown on the dikes in low input system, whereas mainly papaya was grown on the dikes in high input system and the dike outputs in the systems were not significantly different. Crop, fish, livestock, poultry and vegetable were the major enterprises for integration in the studied farms. The nutrient flows among the enterprises in rich farms were more in comparison to poor farms. In the third phase, attempts were made to improve the existing pond-dike system through some technological interventions for better nutrient management The first trial was conducted in 12 farmers' ponds to assess the performance of 3 different aquaculture management systems: system 1 - existing carp polyculture, system 2 - existing carp polyculture + 10% mono-sex male GIFT tilapia, and system 3 - existing carp polyculture + 10% mono- sex male GIFT tilapia + increased feed and fertilizers for the improvement of the existing carp polyculture. The highest production (5,312 kg ha1) was obtained in system 3, which was significantly higher (P < 0.05) from that of the production from systems 2 (2,083 kg ha1) and 1 (1,848 kg ha1). The second trial was carried out to evaluate suitability of pond mud for growing a common leafy vegetable, red amaranth (Amaranthus tricolor). Muds from ponds belonging to the three systems were mixed with the dike plot soils at the rate of 33% (mix 1), 66% (mix 2) and 100% (mix 3). The yields of 4.19 t ha', 1.15 t ha1 and 7.4 t ha1 were obtained in systems 1, 2 and 3, respectively, which were significantly different (P<0.05). Significant difference (P<0.05) was also observed in the yields of 1.76 t ha11, 4.33 t ha1 and 6.65 t ha in mix 1, 2 and 3. The highest yield, 12.13 t ha1 was obtained by the interaction of system 3 and mix 3 and the lowest yield, of 0.57 t ha1 by the interaction of system 2 and mix 1.