The growth of Nile tilapia in cage in the present experiment was relatively low, with daily weight gain of 0.6 g/fish/day, compared with other integrated cage-cum-pond system (1.0 g/fish/day, Yadav et al. 2007; 1.0 g/fish/day, Shrestha, 2000c; and 4.0 - 4.6 g/fish/day, Yi et al., 1996). Similarly, the daily weight gain of African catfish in the present experiment was 0.8 - 1.0 g/fish/day, which was lower than in outdoor cement tanks (1.1 - 1.7 g/fish/day, Yi et al., 2004; and 1.7 - 1.9 g/fish/day, Long and Yi, 2004), an integrated pen-cum-pond system (2.5 - 2.6 g/fish/day, Yi et al., 2003), and integrated cage-cum-pond system (2.1 - 2.2 g/fish/day, Lin and Diana, 1995; and 1.3 g/fish/day, Shrestha et al., 2009), but higher than those in two other integrated cage-cum-pond systems (0.7 g/fish/day, Uddomkarn, 1989; and 0.8 - 0.9 g/fish/day, Ye, 1991). The higher FCR of tilapia and catfish were caused by higher mortality. The possible reason for lower survival was lower early morning DO and prolonged duration of low DO levels in the case of Nile tilapia, and small stock size and cannibalism in the case of African catfish.
The extrapolated carp yield in the control ponds in the present study (3.3 t/ha/year) was comparable to the yield of semi-intensive carp polyculture system of Nepal (3.3 t/ha/year, DoFD, 2009). The combined net yields of tilapia, catfish and carps in tilapia-carps (7.3 t/ha/year) and tilapia-catfish-carp integration system (8.8 t/ha/year) in the present study was lower than the production of 8 - 15 t/ha/year in other cage-cum-pond integration systems (Yi et al., 1996; Yi, 1997; Yi and Lin, 2001; Shrestha 2002).
Both the control and cage treatment produced positive net returns ranging from 292,125 NRs/ha in the control, and 586,662 to 786,135 NRs/ha in the cage treatment. There was also a significant increase in net returns for the integrated cage-cum-pond culture system as compared to the semi-intensive culture of carps alone. However, in the present study, on the basis of profitability, the tilapia-carp integration system is the best. Small farmers having a single pond can produce more fish for sale from cages and carps without feeding in ponds for home consumption as well as for sale. This increased production per unit area as well as income by 2 times than the normal pond culture of carps in Nepal.
The cage-cum-pond integrated system was developed to integrate intensive feeding in cages and semi-intensive fertilization in open ponds, with fertilizer derived from cage wastes. The similar growth rate of the carps in cage-cum-pond integration compared to the fertilized pond without cages indicated that the nutrients released by the cage are sufficient for production of carps in open pond. This experiment demonstrated that the cage-cum-pond integration with Nile tilapia in cage and carps in open pond is one of the best technologies to increase production and profitability for small farmers.